LUZ AZLOR, ANNA PIIL DAMM AND

MARIE LOUISE SCHULTZ-NIELSEN

study paper 132 november 2018

LocaL Labour DemanD anD ImmIgrant empLoyment

The ROCKWOOL Foundation Research Unit

Study Paper No. 132

LUZ AZLOR, ANNA PIIL DAMM AND

MARIE LOUISE SCHULTZ-NIELSEN

Copenhagen 2018

LocaL Labour DemanD anD ImmIgrant empLoyment

Local Labour Demand and Immigrant Employment

Study Paper No. 132

Published by:© The ROCKWOOL Foundation Research Unit

Address:The ROCKWOOL Foundation Research UnitNy Kongensgade 61472 København K.

Telephone +45 33 34 48 00E-mail kontakt@rff.dkweb site: https://www.rockwoolfonden.dk/en

November 2018

3 1

Local Labour Demand and Immigrant Employment*

Luz Azlor,† Anna Piil Damm‡ and Marie Louise Schultz-Nielsen§

October 3rd 2018

Abstract: This paper investigates the effect of local labour demand on employment of immigrant

workers. We take into account self-selection into locations by estimating the effects for refugees

who were subject to the Danish Spatial Dispersal Policy from 1999-2010 using full population

Danish administrative registers that contain information on admission class of immigrants. We

identify refugee status without any measurement error. Our findings show that residence in a

municipality with a one percentage point higher employment rate increases the employment rate

of refugees by 0.6-0.7 percentage point (or 2.1%) within the first four years of their stay in

Denmark. We also argue that the local employment rate is a better measure of local labour

demand for refugees than the local unemployment rate.

Keywords: Immigrants, Refugees, Asylum Seekers, Settlement Policies, Employment.

JEL codes: J23, J61, J68, J71

* This paper acknowledges the support from and access to Statistics Denmark provided by the ROCKWOOL Foundation Research Unit. We acknowledge financial support from the ROCKWOOL Foundation Research Unit, TrygFonden and the Department of Economics and Business Economics, Aarhus University. We thank Bente Herbst Bendiksen and Janne Lindblad at the Danish Immigration Service for sharing their internal administrative statistics and knowledge about the Danish Spatial Dispersal Policy 1999-2016 with us. We thank Peter Fredriksson for his comments on an earlier draft of the paper. We also thank Mie Hjortskov Andersen, Drilon Helshani and Villiam Vellev for research assistance. The authors declare that they have no relevant or material financial interests that relate to the research described in this paper. † Department of Economics and Business, Aarhus University, Fuglesangs Allé 4, DK-8210 Aarhus V. Email: lav@econ.au.dk. Present address: NBI, Freetown, Sierra Leone. Email: lazlor@nbiconsultancy.com. ‡ Department of Economics and Business, Aarhus University, Fuglesangs Allé 4, DK-8210 Aarhus V. Email: apd@econ.au.dk. § ROCKWOOL Foundation Research Unit, Ny Kongensgade 6, 1472 København K. Email: mls@rff.dk.

4 2

I. INTRODUCTION

Increasing rates of immigrants over the past decades in Western countries have spurred debates about immigration and integration policies, questioning whether the host economies can successfully integrate immigrants into the labour market (Bauer, Lofstrom and Zimmermann 2000, Dustmann, Vasiljeva and Damm 2018). Recently, two major events have in particular spurred the debate in Europe. The 2004 and 2007 enlargements of the common European labour market, which triggered a massive inflow of labour migrants from Eastern Europe to the old EU countries, and the massive influx of refugees, notably from Syria, to EU countries, which culminated in the fall 2015.

In a world with large cross-country productivity differences, there is a potential for substantial economic gains from immigration, as open borders allow labour to flow towards its best use (Kennan 2013; Bratsberg, Raaum and Røed 2017). Moreover, immigration of labour may alleviate the demographic and fiscal challenges facing European countries with ageing populations (Storesletten 2000; Bratsberg et al. 2017).

Although employment and earnings of immigrants increase with years spent in the host country (Chiswick 1978; Borjas 1985; LaLonde and Topel 1992; Dustmann 1993; Borjas 1995; Lubotsky 2007; Algan, Dustmann, Glitz and Manning 2010; Sarvimäki 2011, Dustmann and Görlach 2015), studies have documented substantial employment and earnings disparities between immigrants and natives, and for some groups of immigrants, in particular non-labour migrants, the immigrant-native employment gap remains large even 7-10 years after immigration (Edin, Fredriksson and Åslund 2003; Damm 2009; Damm and Rosholm 2010; Bratsberg et al. 2017; Schultz-Nielsen 2017).1

A substantial part of the literature on immigrant employment has investigated the importance of supply-side factors such as admission class, skills acquired in the host country, potential work experience and language ability (Chiswick 1978, Borjas 1985, Borjas 1995, Husted et al. 2001, Cortes 2004, Algan, Dustmann, Glitz and Manning 2010, Dustmann, Glitz and Vogel 2010, Schultz-Nielsen 2016). Since the level of employment is given as the equilibrium between labour supply and labour demand, local labour demand affects employment. Immigrants may in fact be more sensitive to changes in local labour market conditions than natives.2 This may be the case for at least four reasons. First, immigrants are overrepresented in low skilled jobs (Smith et al. 2003; Edin, Fredriksson and Åslund 2004) which fuels instability, in part due to skill- or routine-biased technological change (Katz and Murphy 1992; Berman, Bound and

1 The non-Western immigrant-native employment gaps are particularly large in the Nordic countries, partly due to the high labour force participation rates of natives (including that of women) on which the Nordic welfare models rely (Bratsberg et al. 2014; Bratsberg et al. 2017; Schultz-Nielsen 2017; Åslund, Forslund and Liljeberg 2017; Sarvimäki 2017). 2 See Hoynes (2000) for empirical evidence for the U.S.

3

Machin 1998; Card and DiNardo 2002; Moore and Ranjan 2005; Goos, Manning and Salomons 2014) and task offshoring (Grossman and Rossi-Hansberg 2008; Goos, Manning and Salomons 2014).3 Second, employers may discriminate immigrant applicants.4 Third, in many countries firms use the last-in-first-out (LIFO) principle in downsizing and immigrants are likely to be overrepresented in the group of recent hires (Bratsberg et al. 2017; Åslund et al. 2017). Fourth, job-referral networks of recent immigrant cohorts are ethnically stratified.5

The question of how sensitive immigrant employment is to local labour demand conditions is important for several reasons. First, such knowledge will give us an understanding of the extent to which economic growth alone can increase the employment rate of immigrants. Second, such knowledge can be used for optimal design of public employment policies. For instance, during economic recession it may be optimal to increase resources for employment programs for immigrants or low-skilled workers to stimulate local demand for their skills and spend additional resources on training and skill-upgrading programs. Third, such knowledge can be used for optimal design of settlement policies on newly recognized refugees and asylum seekers. In particular, current settlement policies on refugees employed in a number of European countries can be reformed in order to increase the speed of labour market integration of refugees.

With the exception of Åslund and Rooth (2007) and Damm and Rosholm (2010), few studies identify the causal effect of local labour market conditions on immigrant labour market outcomes. This study helps fill the gap. For identification of the effects of local labour demand on immigrant employment we exploit the Danish Spatial Dispersal Policy on Refugees in place since 1999. Our study is the first to exploit the current Danish Spatial Dispersal Policy on Refugees for identification of causal effects of local characteristics on integration of immigrants into the host country society. We estimate the

3 In Denmark, as in many other countries, the largest share of non-Western immigrants work in the service industry (around 32%), which has not been affected by routine-biased technological changes and taskoffshoring to the same extent as the manufacturing industry. Across industries in Denmark over the 2008-2016-period, the manufacturing industry has experienced the largest reduction in the share of workers: 2.2 percentage points, which given its employment share of 13.8% in 2008 corresponds to 15.8% reduction. With an employment share of 14.3% in manufacturing, non-Western immigrants were slightly overrepresented relative to natives in 2008, but with a share of only 10.4% in manufacturing in 2016, non-Western immigrants are underrepresented in manufacturing. (Authors’ own calculations from Danish public employment statistics across industries and immigrant status, http://www.statistikbanken.dk/RAS311). 4 For empirical evidence from correspondence studies of discrimination by ethnic origin, see Riach and Rich (1991), Esmail and Everington (1997), Bertrand and Mullainathan (2004), Carlsson and Rooth (2007). 5 For descriptive evidence for the U.S., see Munshi (2003) and Beaman (2012). For descriptive evidence for Scandinavia, see Edin, Fredriksson and Åslund (2003) and Damm (2009, 2014).

5 2

I. INTRODUCTION

Increasing rates of immigrants over the past decades in Western countries have spurred debates about immigration and integration policies, questioning whether the host economies can successfully integrate immigrants into the labour market (Bauer, Lofstrom and Zimmermann 2000, Dustmann, Vasiljeva and Damm 2018). Recently, two major events have in particular spurred the debate in Europe. The 2004 and 2007 enlargements of the common European labour market, which triggered a massive inflow of labour migrants from Eastern Europe to the old EU countries, and the massive influx of refugees, notably from Syria, to EU countries, which culminated in the fall 2015.

In a world with large cross-country productivity differences, there is a potential for substantial economic gains from immigration, as open borders allow labour to flow towards its best use (Kennan 2013; Bratsberg, Raaum and Røed 2017). Moreover, immigration of labour may alleviate the demographic and fiscal challenges facing European countries with ageing populations (Storesletten 2000; Bratsberg et al. 2017).

Although employment and earnings of immigrants increase with years spent in the host country (Chiswick 1978; Borjas 1985; LaLonde and Topel 1992; Dustmann 1993; Borjas 1995; Lubotsky 2007; Algan, Dustmann, Glitz and Manning 2010; Sarvimäki 2011, Dustmann and Görlach 2015), studies have documented substantial employment and earnings disparities between immigrants and natives, and for some groups of immigrants, in particular non-labour migrants, the immigrant-native employment gap remains large even 7-10 years after immigration (Edin, Fredriksson and Åslund 2003; Damm 2009; Damm and Rosholm 2010; Bratsberg et al. 2017; Schultz-Nielsen 2017).1

A substantial part of the literature on immigrant employment has investigated the importance of supply-side factors such as admission class, skills acquired in the host country, potential work experience and language ability (Chiswick 1978, Borjas 1985, Borjas 1995, Husted et al. 2001, Cortes 2004, Algan, Dustmann, Glitz and Manning 2010, Dustmann, Glitz and Vogel 2010, Schultz-Nielsen 2016). Since the level of employment is given as the equilibrium between labour supply and labour demand, local labour demand affects employment. Immigrants may in fact be more sensitive to changes in local labour market conditions than natives.2 This may be the case for at least four reasons. First, immigrants are overrepresented in low skilled jobs (Smith et al. 2003; Edin, Fredriksson and Åslund 2004) which fuels instability, in part due to skill- or routine-biased technological change (Katz and Murphy 1992; Berman, Bound and

1 The non-Western immigrant-native employment gaps are particularly large in the Nordic countries, partly due to the high labour force participation rates of natives (including that of women) on which the Nordic welfare models rely (Bratsberg et al. 2014; Bratsberg et al. 2017; Schultz-Nielsen 2017; Åslund, Forslund and Liljeberg 2017; Sarvimäki 2017). 2 See Hoynes (2000) for empirical evidence for the U.S.

3

Machin 1998; Card and DiNardo 2002; Moore and Ranjan 2005; Goos, Manning and Salomons 2014) and task offshoring (Grossman and Rossi-Hansberg 2008; Goos, Manning and Salomons 2014).3 Second, employers may discriminate immigrant applicants.4 Third, in many countries firms use the last-in-first-out (LIFO) principle in downsizing and immigrants are likely to be overrepresented in the group of recent hires (Bratsberg et al. 2017; Åslund et al. 2017). Fourth, job-referral networks of recent immigrant cohorts are ethnically stratified.5

The question of how sensitive immigrant employment is to local labour demand conditions is important for several reasons. First, such knowledge will give us an understanding of the extent to which economic growth alone can increase the employment rate of immigrants. Second, such knowledge can be used for optimal design of public employment policies. For instance, during economic recession it may be optimal to increase resources for employment programs for immigrants or low-skilled workers to stimulate local demand for their skills and spend additional resources on training and skill-upgrading programs. Third, such knowledge can be used for optimal design of settlement policies on newly recognized refugees and asylum seekers. In particular, current settlement policies on refugees employed in a number of European countries can be reformed in order to increase the speed of labour market integration of refugees.

With the exception of Åslund and Rooth (2007) and Damm and Rosholm (2010), few studies identify the causal effect of local labour market conditions on immigrant labour market outcomes. This study helps fill the gap. For identification of the effects of local labour demand on immigrant employment we exploit the Danish Spatial Dispersal Policy on Refugees in place since 1999. Our study is the first to exploit the current Danish Spatial Dispersal Policy on Refugees for identification of causal effects of local characteristics on integration of immigrants into the host country society. We estimate the

3 In Denmark, as in many other countries, the largest share of non-Western immigrants work in the service industry (around 32%), which has not been affected by routine-biased technological changes and taskoffshoring to the same extent as the manufacturing industry. Across industries in Denmark over the 2008-2016-period, the manufacturing industry has experienced the largest reduction in the share of workers: 2.2 percentage points, which given its employment share of 13.8% in 2008 corresponds to 15.8% reduction. With an employment share of 14.3% in manufacturing, non-Western immigrants were slightly overrepresented relative to natives in 2008, but with a share of only 10.4% in manufacturing in 2016, non-Western immigrants are underrepresented in manufacturing. (Authors’ own calculations from Danish public employment statistics across industries and immigrant status, http://www.statistikbanken.dk/RAS311). 4 For empirical evidence from correspondence studies of discrimination by ethnic origin, see Riach and Rich (1991), Esmail and Everington (1997), Bertrand and Mullainathan (2004), Carlsson and Rooth (2007). 5 For descriptive evidence for the U.S., see Munshi (2003) and Beaman (2012). For descriptive evidence for Scandinavia, see Edin, Fredriksson and Åslund (2003) and Damm (2009, 2014).

6 4

effects of local labour market conditions on immigrant employment for refugees who were subject to the policy. An important strength of our paper relative to the before-mentioned previous studies is that we identify refugee status without any measurement error using Danish administrative registers from 1999. Our study hereby addresses an important concern in previous studies that the estimated effects of local labour market conditions are biased due to measurement error stemming from potential use of a contaminated sample.

The structure of our paper is as follows. Section II briefly reviews the existing literature on the link between local labour market conditions and immigrant employment. Section III gives the institutional background. In Section IV we provide our methodological considerations and set up our empirical model. Then follows a description of our data in Section V, and a presentation of our empirical results in Section VI. Section VII provides our concluding remarks.

II. LITERATURE REVIEW

Using administrative data for a 1% sample of participants in the Aid to Families with Dependent Children (AFDC) in California, Hoynes (2000) estimates the effects of different measures of local labour demand on transitions off welfare and transitions back to welfare. She estimates discrete duration models, controlling for demographic and neighbourhood characteristics, duration effects, county fixed effects, time effects and county-specific time trends. The control for county of residence means control for time-constant county characteristics. In other words, the estimation only uses within-county variation for identification of the effects. The identification strategy relies on weaker identifying assumptions than most studies on the effects of local labour market conditions on individual labour market outcomes. The results show that labour market fluctuations are important determinants of both leaving welfare and recidivism into welfare and that Hispanics, Blacks, residents of urban areas and unemployed parent recipients are more sensitive to changes in labour market conditions while whites and teen parents are less sensitive. Another finding is that models that measure labour market conditions using employment-based measures perform better than the model that measures labour market conditions using unemployment rates.

Åslund and Rooth (2007) investigate the effects of initial labour market conditions as measured by the local unemployment rate on immigrant earnings and employment up to 11 years since migration. They address potential selective immigration and location sorting of immigrants by exploiting the whole of Sweden’s settlement program for newly recognized refugees and asylum seekers. Using administrative register data for immigrants from refugee-sending countries who immigrated during the period 1987-91 and controlling for demographic characteristics and level of education, source country,

5

calendar time and years since immigration, they find that initial high local unemployment lowers earnings and employment up to ten years after immigration.

Exploiting the Danish Spatial Dispersal Policy on Refugees run from 1986 until 1998 to address potential selective immigration and location sorting of immigrants, Damm and Rosholm (2010) estimate a bivariate Mixed Proportional Harzard model to obtain the effects of local labour market and housing conditions on the hazard rate into employment and the hazard rate of moving out of the municipality of assignment. The model controls for demographic characteristics and educational level at immigration, calendar time and ethnic group size in the host country using two specifications: without and with county of residence fixed effects. The results show that a higher local unemployment rate decreases the hazard rate into employment; inclusion of county fixed effects renders the estimate insignificant, but the sign is unchanged. The results show further that residence in a large municipality and in a municipality with many immigrants decreases the hazard rate into employment. The number of co-nationals in the municipality has a positive but insignificant effect on the hazard rate into employment. This result is somewhat surprising since Edin et al. (2003) and Damm (2009) find positive effects of the ethnic enclave size on earnings of refugees, but Damm (2009, 2014) also finds an insignificant effect of the ethnic enclave size on the employment of refugees. The joint finding of a negative effect of the presence of immigrants and the local population size of the hazard into employment lends empirical support to government policies that settle refugees in areas outside the immigrant-dense cities in order to promote economic assimilation of refugees.

III. INSTITUTIONAL BACKGROUND

Denmark has had spatial dispersal policies for refugees and asylum seekers who had their applications approved since 1986 (Damm 2005). Henceforth, we refer to such recognized refugees and asylum seekers as “refugees”. The purpose has been to disperse refugees equally across Danish regions and municipalities to ensure that the integration task is shared equitably across the country and to avoid localizing newly arrived refugees in areas were the concentration of foreign nationals is already high, which could potentially hinder refugees’ introduction to the Danish language or society in general.

Until 1998 it was ’The Danish Refugee Council’ (DRC) that organized the placement of the spatial dispersal of the refugees and were in charge of the 18 months long introduction program that included training in Danish language, culture and job training. The goal of the placement was to distribute refugees equally in proportion to the population size. However, to promote ethnic networks, refugees were spatially dispersed in clusters with fellow countrymen. Although the refugees were encouraged to stay in the

7 4

effects of local labour market conditions on immigrant employment for refugees who were subject to the policy. An important strength of our paper relative to the before-mentioned previous studies is that we identify refugee status without any measurement error using Danish administrative registers from 1999. Our study hereby addresses an important concern in previous studies that the estimated effects of local labour market conditions are biased due to measurement error stemming from potential use of a contaminated sample.

The structure of our paper is as follows. Section II briefly reviews the existing literature on the link between local labour market conditions and immigrant employment. Section III gives the institutional background. In Section IV we provide our methodological considerations and set up our empirical model. Then follows a description of our data in Section V, and a presentation of our empirical results in Section VI. Section VII provides our concluding remarks.

II. LITERATURE REVIEW

Using administrative data for a 1% sample of participants in the Aid to Families with Dependent Children (AFDC) in California, Hoynes (2000) estimates the effects of different measures of local labour demand on transitions off welfare and transitions back to welfare. She estimates discrete duration models, controlling for demographic and neighbourhood characteristics, duration effects, county fixed effects, time effects and county-specific time trends. The control for county of residence means control for time-constant county characteristics. In other words, the estimation only uses within-county variation for identification of the effects. The identification strategy relies on weaker identifying assumptions than most studies on the effects of local labour market conditions on individual labour market outcomes. The results show that labour market fluctuations are important determinants of both leaving welfare and recidivism into welfare and that Hispanics, Blacks, residents of urban areas and unemployed parent recipients are more sensitive to changes in labour market conditions while whites and teen parents are less sensitive. Another finding is that models that measure labour market conditions using employment-based measures perform better than the model that measures labour market conditions using unemployment rates.

Åslund and Rooth (2007) investigate the effects of initial labour market conditions as measured by the local unemployment rate on immigrant earnings and employment up to 11 years since migration. They address potential selective immigration and location sorting of immigrants by exploiting the whole of Sweden’s settlement program for newly recognized refugees and asylum seekers. Using administrative register data for immigrants from refugee-sending countries who immigrated during the period 1987-91 and controlling for demographic characteristics and level of education, source country,

5

calendar time and years since immigration, they find that initial high local unemployment lowers earnings and employment up to ten years after immigration.

Exploiting the Danish Spatial Dispersal Policy on Refugees run from 1986 until 1998 to address potential selective immigration and location sorting of immigrants, Damm and Rosholm (2010) estimate a bivariate Mixed Proportional Harzard model to obtain the effects of local labour market and housing conditions on the hazard rate into employment and the hazard rate of moving out of the municipality of assignment. The model controls for demographic characteristics and educational level at immigration, calendar time and ethnic group size in the host country using two specifications: without and with county of residence fixed effects. The results show that a higher local unemployment rate decreases the hazard rate into employment; inclusion of county fixed effects renders the estimate insignificant, but the sign is unchanged. The results show further that residence in a large municipality and in a municipality with many immigrants decreases the hazard rate into employment. The number of co-nationals in the municipality has a positive but insignificant effect on the hazard rate into employment. This result is somewhat surprising since Edin et al. (2003) and Damm (2009) find positive effects of the ethnic enclave size on earnings of refugees, but Damm (2009, 2014) also finds an insignificant effect of the ethnic enclave size on the employment of refugees. The joint finding of a negative effect of the presence of immigrants and the local population size of the hazard into employment lends empirical support to government policies that settle refugees in areas outside the immigrant-dense cities in order to promote economic assimilation of refugees.

III. INSTITUTIONAL BACKGROUND

Denmark has had spatial dispersal policies for refugees and asylum seekers who had their applications approved since 1986 (Damm 2005). Henceforth, we refer to such recognized refugees and asylum seekers as “refugees”. The purpose has been to disperse refugees equally across Danish regions and municipalities to ensure that the integration task is shared equitably across the country and to avoid localizing newly arrived refugees in areas were the concentration of foreign nationals is already high, which could potentially hinder refugees’ introduction to the Danish language or society in general.

Until 1998 it was ’The Danish Refugee Council’ (DRC) that organized the placement of the spatial dispersal of the refugees and were in charge of the 18 months long introduction program that included training in Danish language, culture and job training. The goal of the placement was to distribute refugees equally in proportion to the population size. However, to promote ethnic networks, refugees were spatially dispersed in clusters with fellow countrymen. Although the refugees were encouraged to stay in the

8 6

same municipality of assignment, the social benefits were not conditional on the refugees staying there (Ibid 2005).

III.A. The Danish Spatial Dispersal Policy 1999-2016

With the Danish Parliament’s enactment of the ‘Integration Law’, the introduction program in 1999 was prolonged to 3 years and the responsibility for both this program and the spatial dispersal policy reception was handed over to the municipalities hosting the refugees. The new legislation further tied receipt of welfare benefits to residing in the assigned community. Before 1999 refugees often stayed in the municipality of assignment in the years after the initial settlement, and this tendency was strengthened further after 1999 (Nielsen and Jensen 2006).

The legal basis for the spatial dispersal policy is stipulated in chapter 3 of the ‘Integration Law’. It specifies that The Danish Immigration Service (DIS) each year shall make a forecast (called ‘Landstallet’) of the overall number of refugees that are expected to arrive in the following calendar year. Based on this forecast, allocation of refugees, in the first place, to regions and then municipalities is settled in agreement between these local authorities.6 The allocation is based on a quota-system that is calculated using the region’s/municipality’s share of the total population and the share of foreigners in the region/municipality. This calculation method has remained the same from 1999 to 2016.7 Nevertheless, the annual variation in the number of arriving refugees has hindered the forecasts of DIS. Therefore, DIS adjusts the quotas if the actual number of refugees that arrives within a year differs substantially from the expected number.

It is the responsibility of DIS to refer each refugee to a municipality that has not yet met its yearly quota. Only under extraordinary circumstances will a refugee be referred to a municipality with a full quota. During the asylum process a caseworker from DIS has a meeting with the asylum seeker, first and foremost to secure the correct identity of the asylum seeker and other questions related to the asylum case, but the asylum seekers’ wishes regarding settlement, in case he/she will be granted residence, may also be considered (DIS-interview8).9 Close family already living in Denmark is primarily

6 If the local authorities do not reach an agreement DIS determines the allocation based on calculated quotas. 7 From 1999 to July 2016 the share of foreigners includes foreign nationals, expect those from Nordic countries, EU/EEA. After July 2016 the definition of foreigners used for the quota-calculation is slightly changed and placement should include employment considerations (“Order/BEK # 980 of 28/06/2016”). 8 The authors conducted an interview about the administration of the Danish Spatial Dispersal Policy on the January 18th, 2017 with Bente Herbst Bendiksen and Janne Lindblad at the Danish Immigration Service (DIS). 9 This feature of the Danish Spatial Dispersal Policy on Refugees in place since Jan. 1st 1999 was not part of the first Danish Spatial Dispersal Policy of Refugees in place from 1986-1998. Under the first policy,

7

considered and spouses and children are always settled in the same municipality as the first arrived family member. But other conditions that can be taken into consideration are: nationality and thereby the refugee’s possibility of creating a network with countrymen, educational qualifications and special (medical) treatment (“Order/BEK # 630 of 25/08/1998”). Municipalities on their side may also have wishes related to the above-mentioned characteristics of the refugees they will receive. However, in general educational qualifications from the refugees’ home countries are not easily transferred to the Danish labour market and the municipalities’ desire for special educational groups are modest, just like health-related problems need to be very severe (or even terminal) to be decisive for the placement (DIS-interview).

Hence, the yearly assigned municipality quotas are the core element in the Danish Spatial Dispersal Policy in place from 1999 until 2016. At the beginning of each calendar year settlement of refugees is possible for DIS in all municipalities (with a quota), but as the months pass and more refugees are granted residence, the municipal quotas become filled. So, refugees’ possible preferences regarding the settlement can be more easily accommodated by DIS if the refugee obtains a residence permit earlier in the calendar year rather than later during the same calendar year. Because if a refugee wishes to go to a municipality that has already fulfilled its quota for the year, he or she will instead be settled in one of the municipalities that has still not reached its quota. Importantly, this aspect of the refugee settlement policy is a novel finding of our interview with DIS and has not been discussed in public. Besides, the date at which a refugee is assigned to a municipality can be considered outside the control of the refugee himself, since municipal assignment takes place shortly after receipt of asylum and since asylum seekers wait for months (or even years) to obtain a Danish residence permit. Thus, investigating the effects of local labour market conditions for the subsample of refugees assigned to a municipality in the later months of the year resembles a field experiment.

IV. METHODOLOGICAL CONSIDERATIONS AND EMPIRICAL MODEL

IV.A. Methodological considerations

The main challenge in identification of the effects of local labour demand conditions on immigrant employment arises because immigrants may sort into locations in terms of individual characteristics which are unobserved by researchers. Previous research has shown that this is indeed the case. Among refugees subject to a spatial dispersal program at the time of receipt of asylum, individuals who subsequently moved into ethnic

placement officers did not interview newly recognized asylum seekers and they assigned refugees to locations with little or no regard of location wishes (see Damm 2014; Damm and Dustmann 2014).

9 7

considered and spouses and children are always settled in the same municipality as the first arrived family member. But other conditions that can be taken into consideration are: nationality and thereby the refugee’s possibility of creating a network with countrymen, educational qualifications and special (medical) treatment (“Order/BEK # 630 of 25/08/1998”). Municipalities on their side may also have wishes related to the above-mentioned characteristics of the refugees they will receive. However, in general educational qualifications from the refugees’ home countries are not easily transferred to the Danish labour market and the municipalities’ desire for special educational groups are modest, just like health-related problems need to be very severe (or even terminal) to be decisive for the placement (DIS-interview).

Hence, the yearly assigned municipality quotas are the core element in the Danish Spatial Dispersal Policy in place from 1999 until 2016. At the beginning of each calendar year settlement of refugees is possible for DIS in all municipalities (with a quota), but as the months pass and more refugees are granted residence, the municipal quotas become filled. So, refugees’ possible preferences regarding the settlement can be more easily accommodated by DIS if the refugee obtains a residence permit earlier in the calendar year rather than later during the same calendar year. Because if a refugee wishes to go to a municipality that has already fulfilled its quota for the year, he or she will instead be settled in one of the municipalities that has still not reached its quota. Importantly, this aspect of the refugee settlement policy is a novel finding of our interview with DIS and has not been discussed in public. Besides, the date at which a refugee is assigned to a municipality can be considered outside the control of the refugee himself, since municipal assignment takes place shortly after receipt of asylum and since asylum seekers wait for months (or even years) to obtain a Danish residence permit. Thus, investigating the effects of local labour market conditions for the subsample of refugees assigned to a municipality in the later months of the year resembles a field experiment.

IV. METHODOLOGICAL CONSIDERATIONS AND EMPIRICAL MODEL

IV.A. Methodological considerations

The main challenge in identification of the effects of local labour demand conditions on immigrant employment arises because immigrants may sort into locations in terms of individual characteristics which are unobserved by researchers. Previous research has shown that this is indeed the case. Among refugees subject to a spatial dispersal program at the time of receipt of asylum, individuals who subsequently moved into ethnic

placement officers did not interview newly recognized asylum seekers and they assigned refugees to locations with little or no regard of location wishes (see Damm 2014; Damm and Dustmann 2014).

10 8

enclaves were negatively selected in terms of individual unobservable characteristics, for instance, English language proficiency (Edin et al. 2003; Damm 2009). Therefore, use of observational data for estimation of the effects of local labour demand conditions on immigrant employment will result in biased results due to omitted variables. Instead, we estimate the effects for the population of refugees who were assigned to housing across municipalities in Denmark upon receipt of asylum after the first municipal quota had been filled. In our regressions we condition on the characteristics of the household head which were observed by DIS at the time of assignment.

Local labour demand can be measured in different ways. The local unemployment rate is a common measure which reflects excess supply of labour at the minimum wage (according to the neoclassical theory of the firm). The local unemployment rate is also negatively correlated with labour market tightness defined as the number of job vacancies relative to the number of unemployed (according to job search theory). By definition, the unemployment rate is also negatively correlated with the labour force participation rate. In a situation with excess supply of labour, long-term unemployed workers may leave the work force as discouraged workers and re-enter the work force again when the local labour demand increases again. Thus, both the nominator and denominator in the unemployment rate would change, leaving the unemployment rate relatively unaffected by the number of discouraged workers. Besides, long-term unemployed who are no longer entitled to unemployment insurance benefits have little financial incentive to stay in the work force. Unskilled as well as skilled workers with obsolete skills or, in case of immigrant workers, not easily transferable skills from the source country, are likely to be overrepresented among those long-term unemployed and discouraged workers. In fact, the share of discouraged workers is non-negligible according to the OECD. Therefore, the local unemployment rate may not be an accurate measure of excess supply of workers in business cycles downturns.

Alternatively, the local employment rate (defined as the number of employed relative to the population in the working ages) may be a better measure of local labour demand as only the nominator is affected by the number of discouraged workers.10 In other words, the employment rate varies with the number of discouraged workers.

Our analyses concern local labour demand for a particular group of workers, namely immigrants from non-Western countries, who may bring skills that are not easy to transfer to the host country’s labour market and may not be proficient in the host country’s language. For this reason, local demand for their labour may be better

10 Employment is an equilibrium outcome determined by the intersection between labour demand and labour supply. Since the number of discouraged workers affects the labour supply, the level of employment reflects the number of discouraged workers.

9

measured by the local unemployment rate and employment rate among non-Western immigrants, which we include as measures of local labour demand in our analyses.

The final measure of local labour demand we use is the net employment growth, which is equal to the difference between job creation and job destruction. We calculate it as the annual change in the number of employed individuals relative to the number of employed individuals last year.11 Similar to the employment rate, it varies with the number of discouraged workers. But in contrast to the unemployment and employment rates, it does not measure labour demand relative to the (potential) supply as measured by the (potential) size of the labour force.

Identification of the effects of local labour demand on immigrant employment requires control for correlated effects, that is, other characteristics of the local labour market which are correlated with both the local labour demand and individual employment, e.g. city size, job search networks and commuting costs. In the baseline specification we control for the two municipality characteristics which are used by DIS to determine the annual municipal quota for the following year in combination with the expected number of refugees. These two municipality characteristics are municipality size as measured by the share of the Danish population living in the municipality and the non-Western immigrant share. According to the formula used to determine the municipal quota of the expected number of new refugees in the following year, municipalities with a larger population share and larger non-Western immigrant share, receive a disproportionate share of the expected number of new refugees. In robustness checks we include additional controls: i) the co-national share as a measure of job search networks, ii) alternative measures of commuting costs: commuting time to the centre of the commuting area using public transportation, commuting time to the centre of the commuting area by car, distance to the centre of the commuting area in kilometres, and iii) commuting area fixed effects.

IV.B. Empirical model

We use three related empirical models to analyse the importance of local labour market conditions for immigrants’ employment success. Our first basic model describes the association between immigrant’s employment, local labour market conditions, personal characteristics and municipality characteristics within a given year since assignment:

(1) 𝑌𝑌𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖(𝑡𝑡+𝑠𝑠) = 𝛼𝛼1𝐿𝐿𝐿𝐿𝐿𝐿𝑖𝑖(𝑡𝑡+𝑠𝑠) + 𝛽𝛽1𝑋𝑋𝑖𝑖𝑡𝑡 + 𝛾𝛾1𝜈𝜈𝑖𝑖(𝑡𝑡+𝑠𝑠) + 𝛿𝛿𝑖𝑖 + 𝛿𝛿𝑡𝑡 + 𝛿𝛿𝑖𝑖 + 𝜀𝜀𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖(𝑡𝑡+𝑠𝑠)

where the subscripts denote i: individual, j: current municipality of residence, c: country of origin, t: year of municipal assignment, m: month of residence assignment, s: years

11 This measure of local labour market conditions is inspired by Hoynes (2000).

11 8

enclaves were negatively selected in terms of individual unobservable characteristics, for instance, English language proficiency (Edin et al. 2003; Damm 2009). Therefore, use of observational data for estimation of the effects of local labour demand conditions on immigrant employment will result in biased results due to omitted variables. Instead, we estimate the effects for the population of refugees who were assigned to housing across municipalities in Denmark upon receipt of asylum after the first municipal quota had been filled. In our regressions we condition on the characteristics of the household head which were observed by DIS at the time of assignment.

Local labour demand can be measured in different ways. The local unemployment rate is a common measure which reflects excess supply of labour at the minimum wage (according to the neoclassical theory of the firm). The local unemployment rate is also negatively correlated with labour market tightness defined as the number of job vacancies relative to the number of unemployed (according to job search theory). By definition, the unemployment rate is also negatively correlated with the labour force participation rate. In a situation with excess supply of labour, long-term unemployed workers may leave the work force as discouraged workers and re-enter the work force again when the local labour demand increases again. Thus, both the nominator and denominator in the unemployment rate would change, leaving the unemployment rate relatively unaffected by the number of discouraged workers. Besides, long-term unemployed who are no longer entitled to unemployment insurance benefits have little financial incentive to stay in the work force. Unskilled as well as skilled workers with obsolete skills or, in case of immigrant workers, not easily transferable skills from the source country, are likely to be overrepresented among those long-term unemployed and discouraged workers. In fact, the share of discouraged workers is non-negligible according to the OECD. Therefore, the local unemployment rate may not be an accurate measure of excess supply of workers in business cycles downturns.

Alternatively, the local employment rate (defined as the number of employed relative to the population in the working ages) may be a better measure of local labour demand as only the nominator is affected by the number of discouraged workers.10 In other words, the employment rate varies with the number of discouraged workers.

Our analyses concern local labour demand for a particular group of workers, namely immigrants from non-Western countries, who may bring skills that are not easy to transfer to the host country’s labour market and may not be proficient in the host country’s language. For this reason, local demand for their labour may be better

10 Employment is an equilibrium outcome determined by the intersection between labour demand and labour supply. Since the number of discouraged workers affects the labour supply, the level of employment reflects the number of discouraged workers.

9

measured by the local unemployment rate and employment rate among non-Western immigrants, which we include as measures of local labour demand in our analyses.

The final measure of local labour demand we use is the net employment growth, which is equal to the difference between job creation and job destruction. We calculate it as the annual change in the number of employed individuals relative to the number of employed individuals last year.11 Similar to the employment rate, it varies with the number of discouraged workers. But in contrast to the unemployment and employment rates, it does not measure labour demand relative to the (potential) supply as measured by the (potential) size of the labour force.

Identification of the effects of local labour demand on immigrant employment requires control for correlated effects, that is, other characteristics of the local labour market which are correlated with both the local labour demand and individual employment, e.g. city size, job search networks and commuting costs. In the baseline specification we control for the two municipality characteristics which are used by DIS to determine the annual municipal quota for the following year in combination with the expected number of refugees. These two municipality characteristics are municipality size as measured by the share of the Danish population living in the municipality and the non-Western immigrant share. According to the formula used to determine the municipal quota of the expected number of new refugees in the following year, municipalities with a larger population share and larger non-Western immigrant share, receive a disproportionate share of the expected number of new refugees. In robustness checks we include additional controls: i) the co-national share as a measure of job search networks, ii) alternative measures of commuting costs: commuting time to the centre of the commuting area using public transportation, commuting time to the centre of the commuting area by car, distance to the centre of the commuting area in kilometres, and iii) commuting area fixed effects.

IV.B. Empirical model

We use three related empirical models to analyse the importance of local labour market conditions for immigrants’ employment success. Our first basic model describes the association between immigrant’s employment, local labour market conditions, personal characteristics and municipality characteristics within a given year since assignment:

(1) 𝑌𝑌𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖(𝑡𝑡+𝑠𝑠) = 𝛼𝛼1𝐿𝐿𝐿𝐿𝐿𝐿𝑖𝑖(𝑡𝑡+𝑠𝑠) + 𝛽𝛽1𝑋𝑋𝑖𝑖𝑡𝑡 + 𝛾𝛾1𝜈𝜈𝑖𝑖(𝑡𝑡+𝑠𝑠) + 𝛿𝛿𝑖𝑖 + 𝛿𝛿𝑡𝑡 + 𝛿𝛿𝑖𝑖 + 𝜀𝜀𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖(𝑡𝑡+𝑠𝑠)

where the subscripts denote i: individual, j: current municipality of residence, c: country of origin, t: year of municipal assignment, m: month of residence assignment, s: years

11 This measure of local labour market conditions is inspired by Hoynes (2000).

12 10

since migration. The dependent variable 𝑌𝑌𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖(𝑡𝑡+𝑠𝑠) is a dummy for being employed in year t+s. The parameter of interest is 𝛼𝛼1that in turn provides an estimate of the effect of five different measures of local labour market conditions: unemployment rate, unemployment rate among non-Western immigrants, employment rate, employment rate among non-Western immigrants and employment growth. 𝑋𝑋𝑖𝑖𝑡𝑡: Personal characteristics at time t, 𝜈𝜈𝑖𝑖(𝑡𝑡+𝑠𝑠) represents municipality characteristics at time t+s: share of total population and immigrant share, while 𝛿𝛿𝑖𝑖 is country of origin fixed effects, just like 𝛿𝛿𝑡𝑡 and 𝛿𝛿𝑖𝑖 are respectively year and month fixed effects. 𝜀𝜀𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖(𝑡𝑡+𝑠𝑠) is the error term. We estimate the model in Eq. (1) by pooled OLS for s=2, 3, 4.12 𝛼𝛼1 will provide a consistent estimate of the effects of the local labour market characteristic under the strong assumption of no self-selection of immigrants into locations and no omitted correlated effects.

To take account of immigrants’ possible self-selection into municipalities we first restrict our sample to refugees, whom are subject to the Danish Dispersal Policy that assigns them to a municipality upon arrival (at time t) and second restrict our sample to refugees arriving after the first municipality quotas have been filled, making refugees’ own preferences for settlement even less likely to influence municipality assignment.

After settlement in the assigned municipality, refugees with certain unobserved time-varying characteristics like host country language proficiency may sort into municipalities with favourable labour market conditions and thereby influence 𝐿𝐿𝐿𝐿𝐿𝐿𝑖𝑖(𝑡𝑡+𝑠𝑠). To take account of this possible self-selection we follow Åslund and Roth (2007) and introduce a second model where we instrument local labour market conditions for each refugee at time t+s, by the value of the conditions at time of assignment t. Henceforth, we will refer to this model as the IV-model:

(2) 𝑌𝑌𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖(𝑡𝑡+𝑠𝑠) = 𝛼𝛼2𝐿𝐿𝐿𝐿𝐿𝐿𝑖𝑖(𝑡𝑡+𝑠𝑠) + 𝛽𝛽2𝑋𝑋𝑖𝑖𝑡𝑡 + 𝛾𝛾2𝜈𝜈𝑖𝑖∗(𝑡𝑡)+𝛿𝛿𝑖𝑖 + 𝛿𝛿𝑡𝑡 + 𝛿𝛿𝑖𝑖 + 𝜇𝜇𝑖𝑖 + 𝑢𝑢𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖(𝑡𝑡+𝑠𝑠)

𝐿𝐿𝐿𝐿𝐿𝐿𝑖𝑖(𝑡𝑡+𝑠𝑠) = 𝜫𝜫′𝑿𝑿1𝑖𝑖𝑡𝑡 + 𝜐𝜐𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖(𝑡𝑡+𝑠𝑠)

where u is the error term in the main equation and the remaining variables and indices in the main equation are the same as in eq. (1), except that we additionally control for an individual random effect μi. To control for unobserved, time-varying individual characteristics which may be correlated with both 𝐿𝐿𝐿𝐿𝐿𝐿𝑖𝑖(𝑡𝑡+𝑠𝑠) and Y, e.g. language skills, we instrument 𝐿𝐿𝐿𝐿𝐿𝐿𝑖𝑖(𝑡𝑡+𝑠𝑠) in the main equation by z, excluded from X. The instrument z must satisfy two conditions: z should be i) uncorrelated with the error term in the main

12 Since the dependent variable is a dummy variable, in fact we estimate linear probability models. Our model specification uses dummy variables as control variables in order to satisfy the requirements of a saturated model. The saturated model with a discrete outcome will identify identical coefficient estimates and standard errors to a logit/probit model (Angrist, 2001).

11

equation u (the exclusion restriction) and ii) a strong predictor of 𝐿𝐿𝐿𝐿𝐿𝐿𝑖𝑖(𝑡𝑡+𝑠𝑠). In other words, the instrument z must only affect Y through 𝐿𝐿𝐿𝐿𝐿𝐿𝑖𝑖(𝑡𝑡+𝑠𝑠). Apart from z, X1 includes X, 𝜈𝜈, δc, δt, δm and μi. υ is the error term. We instrument 𝐿𝐿𝐿𝐿𝐿𝐿𝑖𝑖(𝑡𝑡+𝑠𝑠) by the local labour market conditions in the municipality of assignment j* in the year of assignment t, 𝐿𝐿𝐿𝐿𝐿𝐿𝑖𝑖∗(𝑡𝑡). In case of self-selection into municipalities, inclusion of 𝜈𝜈𝑖𝑖(𝑡𝑡+𝑠𝑠) in the main equation of Eq. 2 would result in controlling for endogenous regressors. To avoid instrumenting multiple endogenous regressors, we instead control for correlated effects by controlling for other characteristics of the assigned municipality j* in the year of assignment 𝜈𝜈𝑖𝑖∗(𝑡𝑡). We estimate the IV-model for s=2, 3, 4 by 2SLS. By exploiting the panel dimension of our data we account for individual time-constant unobserved heterogeneity like innate abilities by inclusion of an individual random effect 𝜇𝜇𝑖𝑖 which adds efficiency to our model.13 Since DIS did not observe innate abilities, they did not influence the location assignment decision. Therefore, due to the refugee settlement policy, the identifying condition for the RE-estimator of no correlation between the individual random effect and the independent variables is likely to be met. 𝛼𝛼2 identifies the effect of local labour demand under the following assumptions: (i) refugees in our sample were randomly distributed across municipalities, conditional on the observed personal attributes which were known by DIS and may therefore have affected the assignment to municipality type, (ii) the initial labour market conditions only affects the individual’s current employment status through its impact on current labour market conditions (exclusion restriction), (iii) there are no omitted correlated effects, (iv) the time-constant individual unobserved heterogeneity is uncorrelated with the independent variables. If these assumptions are met, 𝛼𝛼2 is the treatment-on-the treated effect and, in case of homogenous treatment effects, the average treatment effect.

Based on the municipality of assignment we finally propose a reduced form model, where refugee’s employment in year t+s is explained by local labour market conditions and municipality characteristics related to the municipality of assignment at time t:

(3) 𝑌𝑌𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖(𝑡𝑡+𝑠𝑠) = 𝛼𝛼3𝐿𝐿𝐿𝐿𝐿𝐿𝑖𝑖∗(𝑡𝑡) + 𝛽𝛽3𝑋𝑋𝑖𝑖𝑡𝑡 + 𝛾𝛾3𝜈𝜈𝑖𝑖∗(𝑡𝑡) + 𝛿𝛿𝑖𝑖 + 𝛿𝛿𝑡𝑡 + 𝛿𝛿𝑖𝑖 + 𝜇𝜇𝑖𝑖 + 𝜖𝜖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖(𝑡𝑡+𝑠𝑠)

where 𝜖𝜖 is the error term. The advantage of the reduced form model is that it gives consistent estimates even if the exclusion restriction (ii) required for consistency of the IV-model is not satisfied. Suppose that labour demand is measured by the local employment rate. Then 𝛼𝛼3 identifies the intent-to-treat estimate of assignment to a municipality with a one percentage point higher employment rate on individual employment under the above-mentioned assumptions (i), (iii) and (iv). Since the

13 It is not possible to include an individual fixed effect instead of a random effect, because for any given refugee, there is only one year of arrival and hence no time-variation in the instrument in the 2SLS or the reduced form.

13 10

since migration. The dependent variable 𝑌𝑌𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖(𝑡𝑡+𝑠𝑠) is a dummy for being employed in year t+s. The parameter of interest is 𝛼𝛼1that in turn provides an estimate of the effect of five different measures of local labour market conditions: unemployment rate, unemployment rate among non-Western immigrants, employment rate, employment rate among non-Western immigrants and employment growth. 𝑋𝑋𝑖𝑖𝑡𝑡: Personal characteristics at time t, 𝜈𝜈𝑖𝑖(𝑡𝑡+𝑠𝑠) represents municipality characteristics at time t+s: share of total population and immigrant share, while 𝛿𝛿𝑖𝑖 is country of origin fixed effects, just like 𝛿𝛿𝑡𝑡 and 𝛿𝛿𝑖𝑖 are respectively year and month fixed effects. 𝜀𝜀𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖(𝑡𝑡+𝑠𝑠) is the error term. We estimate the model in Eq. (1) by pooled OLS for s=2, 3, 4.12 𝛼𝛼1 will provide a consistent estimate of the effects of the local labour market characteristic under the strong assumption of no self-selection of immigrants into locations and no omitted correlated effects.

To take account of immigrants’ possible self-selection into municipalities we first restrict our sample to refugees, whom are subject to the Danish Dispersal Policy that assigns them to a municipality upon arrival (at time t) and second restrict our sample to refugees arriving after the first municipality quotas have been filled, making refugees’ own preferences for settlement even less likely to influence municipality assignment.

After settlement in the assigned municipality, refugees with certain unobserved time-varying characteristics like host country language proficiency may sort into municipalities with favourable labour market conditions and thereby influence 𝐿𝐿𝐿𝐿𝐿𝐿𝑖𝑖(𝑡𝑡+𝑠𝑠). To take account of this possible self-selection we follow Åslund and Roth (2007) and introduce a second model where we instrument local labour market conditions for each refugee at time t+s, by the value of the conditions at time of assignment t. Henceforth, we will refer to this model as the IV-model:

(2) 𝑌𝑌𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖(𝑡𝑡+𝑠𝑠) = 𝛼𝛼2𝐿𝐿𝐿𝐿𝐿𝐿𝑖𝑖(𝑡𝑡+𝑠𝑠) + 𝛽𝛽2𝑋𝑋𝑖𝑖𝑡𝑡 + 𝛾𝛾2𝜈𝜈𝑖𝑖∗(𝑡𝑡)+𝛿𝛿𝑖𝑖 + 𝛿𝛿𝑡𝑡 + 𝛿𝛿𝑖𝑖 + 𝜇𝜇𝑖𝑖 + 𝑢𝑢𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖(𝑡𝑡+𝑠𝑠)

𝐿𝐿𝐿𝐿𝐿𝐿𝑖𝑖(𝑡𝑡+𝑠𝑠) = 𝜫𝜫′𝑿𝑿1𝑖𝑖𝑡𝑡 + 𝜐𝜐𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖(𝑡𝑡+𝑠𝑠)

where u is the error term in the main equation and the remaining variables and indices in the main equation are the same as in eq. (1), except that we additionally control for an individual random effect μi. To control for unobserved, time-varying individual characteristics which may be correlated with both 𝐿𝐿𝐿𝐿𝐿𝐿𝑖𝑖(𝑡𝑡+𝑠𝑠) and Y, e.g. language skills, we instrument 𝐿𝐿𝐿𝐿𝐿𝐿𝑖𝑖(𝑡𝑡+𝑠𝑠) in the main equation by z, excluded from X. The instrument z must satisfy two conditions: z should be i) uncorrelated with the error term in the main

12 Since the dependent variable is a dummy variable, in fact we estimate linear probability models. Our model specification uses dummy variables as control variables in order to satisfy the requirements of a saturated model. The saturated model with a discrete outcome will identify identical coefficient estimates and standard errors to a logit/probit model (Angrist, 2001).

11

equation u (the exclusion restriction) and ii) a strong predictor of 𝐿𝐿𝐿𝐿𝐿𝐿𝑖𝑖(𝑡𝑡+𝑠𝑠). In other words, the instrument z must only affect Y through 𝐿𝐿𝐿𝐿𝐿𝐿𝑖𝑖(𝑡𝑡+𝑠𝑠). Apart from z, X1 includes X, 𝜈𝜈, δc, δt, δm and μi. υ is the error term. We instrument 𝐿𝐿𝐿𝐿𝐿𝐿𝑖𝑖(𝑡𝑡+𝑠𝑠) by the local labour market conditions in the municipality of assignment j* in the year of assignment t, 𝐿𝐿𝐿𝐿𝐿𝐿𝑖𝑖∗(𝑡𝑡). In case of self-selection into municipalities, inclusion of 𝜈𝜈𝑖𝑖(𝑡𝑡+𝑠𝑠) in the main equation of Eq. 2 would result in controlling for endogenous regressors. To avoid instrumenting multiple endogenous regressors, we instead control for correlated effects by controlling for other characteristics of the assigned municipality j* in the year of assignment 𝜈𝜈𝑖𝑖∗(𝑡𝑡). We estimate the IV-model for s=2, 3, 4 by 2SLS. By exploiting the panel dimension of our data we account for individual time-constant unobserved heterogeneity like innate abilities by inclusion of an individual random effect 𝜇𝜇𝑖𝑖 which adds efficiency to our model.13 Since DIS did not observe innate abilities, they did not influence the location assignment decision. Therefore, due to the refugee settlement policy, the identifying condition for the RE-estimator of no correlation between the individual random effect and the independent variables is likely to be met. 𝛼𝛼2 identifies the effect of local labour demand under the following assumptions: (i) refugees in our sample were randomly distributed across municipalities, conditional on the observed personal attributes which were known by DIS and may therefore have affected the assignment to municipality type, (ii) the initial labour market conditions only affects the individual’s current employment status through its impact on current labour market conditions (exclusion restriction), (iii) there are no omitted correlated effects, (iv) the time-constant individual unobserved heterogeneity is uncorrelated with the independent variables. If these assumptions are met, 𝛼𝛼2 is the treatment-on-the treated effect and, in case of homogenous treatment effects, the average treatment effect.

Based on the municipality of assignment we finally propose a reduced form model, where refugee’s employment in year t+s is explained by local labour market conditions and municipality characteristics related to the municipality of assignment at time t:

(3) 𝑌𝑌𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖(𝑡𝑡+𝑠𝑠) = 𝛼𝛼3𝐿𝐿𝐿𝐿𝐿𝐿𝑖𝑖∗(𝑡𝑡) + 𝛽𝛽3𝑋𝑋𝑖𝑖𝑡𝑡 + 𝛾𝛾3𝜈𝜈𝑖𝑖∗(𝑡𝑡) + 𝛿𝛿𝑖𝑖 + 𝛿𝛿𝑡𝑡 + 𝛿𝛿𝑖𝑖 + 𝜇𝜇𝑖𝑖 + 𝜖𝜖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖(𝑡𝑡+𝑠𝑠)

where 𝜖𝜖 is the error term. The advantage of the reduced form model is that it gives consistent estimates even if the exclusion restriction (ii) required for consistency of the IV-model is not satisfied. Suppose that labour demand is measured by the local employment rate. Then 𝛼𝛼3 identifies the intent-to-treat estimate of assignment to a municipality with a one percentage point higher employment rate on individual employment under the above-mentioned assumptions (i), (iii) and (iv). Since the

13 It is not possible to include an individual fixed effect instead of a random effect, because for any given refugee, there is only one year of arrival and hence no time-variation in the instrument in the 2SLS or the reduced form.

14 12

treatment – municipal labour market conditions – varies by municipality, we cluster the standard errors by municipality of assignment.14

V. DATA

The following section provides a description of the raw data and our sample selection criteria. Furthermore, we conduct an initial investigation of the dataset, focusing on the geographical dispersal of refugees across municipalities, the municipal quotas, as well as the labour market attachment of refugees.

V.A. Data sources and sample selection

The empirical analysis presented in this work is based on longitudinal administrative register data from Statistics Denmark for the years between 1999 and 2015. The Danish Immigration Service (DIS) has detailed information on granted residence permits from 1997 onwards, allowing us to perfectly identify refugees for the period of interest. Using a unique person identifier, it is possible to link the data from the DIS with the Danish population register that contains demographic characteristics, e.g. gender, age, residence and other records maintained by Statistics Demark, like the labour market status. We construct three samples of refugees.

The gross sample of refugees comprises the newly arrived adult refugees in the period 1999 to 2015. We focus on refugees arriving as adults in the period 1999 to 2010.15 Individuals that are not recorded in the population registry within a year after they obtain a residence permit are excluded from the gross sample. The gross sample has 12,692 individuals.

Following the steps indicated below, a balanced panel of household heads is constructed. It consists of observations for the subsample of refugees in the gross sample who are observed for at least four years since assignment, who were in their working ages (aged 18 to 59) at the time of asylum and who are household heads. We now explain each of these three selection criteria for extraction of the balanced panel in turn.

14 The treatment also varies by time, but clustering by time as well as municipality would introduce autocorrelation into the error terms. 15 We restrict the sample to refugees arriving before 2011 for two reasons. First, because we wish to extract a panel of refugees whom we can follow in the administrative registers for at least four years since asylum; given that 2015 is our last year of observation, this limits our sample to refugees arriving before 2012. Second, the number of refugees arriving in 2011 is unexpectedly low. As a consequence, location wishes of all refugees arriving in 2011 are likely to have been met; the number of refugees in the 2011 cohort who arrived after the first 10 municipalities had their annual refugee quota filled was close to zero.

13

First, a balanced panel allows us to estimate a panel data model with individual random effects to account for time-constant unobserved individual heterogeneity. The choice of four years stems from our genuine interest to analyse the effects of the initial labour market conditions after the three-year introduction program, without compromising the sample size.

Second, we restrict the balanced panel to refugee households who were in their working ages (aged 18 to 59) at the time they were granted asylum, because our aim is to analyse the integration of refugees into the labour market.16

Finally, we limit the balanced panel of refugees to the first adult member of the family being granted asylum whom we consider the household head. We consider the husband to be the household head if a married couple of refugees is granted asylum on the same date. In principle the assignment to a municipality of the first family member will determine the assignment of close family members to the same municipality, since DIS as described earlier do not split spouses and children even if municipality quotas are filled.17 However, later arrived family members may arrive after the household head has moved away from the municipality of assignment. Inclusion of such family members into our estimation sample would bias our results.

The dataset does not contain exact information on the initial municipality of placement, instead, this is retrieved from the population registers. Particularly, it is possible to trace people’s municipality of residence, determined at the end of each year. We treat the first municipality registered as the municipality of assignment, if the refugee is recorded in the registries the year in which asylum was granted, or on the subsequent year.18

The specific date in which the household head has been assigned to the municipality is key for our identification strategy considering the above-mentioned new findings. The date recorded in the residence permit information, provided by the DIS, is the date at which the individual is recorded in the municipal population register, which happens quite fast after the refugees receive their residence permit.19 With the exception of UN quota refugees (of which Denmark until 2016 invited 500 annually), applicants for

16 Leaving the panel is caused by out-migration, but only 1% of the gross sample out-migrate within the four-year observation period. 17 The household head is defined as the parent, or the father if both parents were granted asylum on the same date, at working age in the family, between 25 and 59 years old. The reason for choosing the male partner is that, as the data descriptive will show, in the context of non-western immigrants, men have a higher labour market attachment than women. 18 Refugees who are neither observed in the population registers in the year of asylum nor the subsequent year are excluded from our balanced sample of refugee household heads. 19 It takes on average 40 days from refugees are permitted residence until they are registered in the municipality population register in the period 2005-2010, for which the calculation has been made (Hvidtfeldt and Schultz-Nielsen 2017).

15 12

treatment – municipal labour market conditions – varies by municipality, we cluster the standard errors by municipality of assignment.14

V. DATA

The following section provides a description of the raw data and our sample selection criteria. Furthermore, we conduct an initial investigation of the dataset, focusing on the geographical dispersal of refugees across municipalities, the municipal quotas, as well as the labour market attachment of refugees.

V.A. Data sources and sample selection

The empirical analysis presented in this work is based on longitudinal administrative register data from Statistics Denmark for the years between 1999 and 2015. The Danish Immigration Service (DIS) has detailed information on granted residence permits from 1997 onwards, allowing us to perfectly identify refugees for the period of interest. Using a unique person identifier, it is possible to link the data from the DIS with the Danish population register that contains demographic characteristics, e.g. gender, age, residence and other records maintained by Statistics Demark, like the labour market status. We construct three samples of refugees.

The gross sample of refugees comprises the newly arrived adult refugees in the period 1999 to 2015. We focus on refugees arriving as adults in the period 1999 to 2010.15 Individuals that are not recorded in the population registry within a year after they obtain a residence permit are excluded from the gross sample. The gross sample has 12,692 individuals.

Following the steps indicated below, a balanced panel of household heads is constructed. It consists of observations for the subsample of refugees in the gross sample who are observed for at least four years since assignment, who were in their working ages (aged 18 to 59) at the time of asylum and who are household heads. We now explain each of these three selection criteria for extraction of the balanced panel in turn.

14 The treatment also varies by time, but clustering by time as well as municipality would introduce autocorrelation into the error terms. 15 We restrict the sample to refugees arriving before 2011 for two reasons. First, because we wish to extract a panel of refugees whom we can follow in the administrative registers for at least four years since asylum; given that 2015 is our last year of observation, this limits our sample to refugees arriving before 2012. Second, the number of refugees arriving in 2011 is unexpectedly low. As a consequence, location wishes of all refugees arriving in 2011 are likely to have been met; the number of refugees in the 2011 cohort who arrived after the first 10 municipalities had their annual refugee quota filled was close to zero.

13

First, a balanced panel allows us to estimate a panel data model with individual random effects to account for time-constant unobserved individual heterogeneity. The choice of four years stems from our genuine interest to analyse the effects of the initial labour market conditions after the three-year introduction program, without compromising the sample size.

Second, we restrict the balanced panel to refugee households who were in their working ages (aged 18 to 59) at the time they were granted asylum, because our aim is to analyse the integration of refugees into the labour market.16

Finally, we limit the balanced panel of refugees to the first adult member of the family being granted asylum whom we consider the household head. We consider the husband to be the household head if a married couple of refugees is granted asylum on the same date. In principle the assignment to a municipality of the first family member will determine the assignment of close family members to the same municipality, since DIS as described earlier do not split spouses and children even if municipality quotas are filled.17 However, later arrived family members may arrive after the household head has moved away from the municipality of assignment. Inclusion of such family members into our estimation sample would bias our results.

The dataset does not contain exact information on the initial municipality of placement, instead, this is retrieved from the population registers. Particularly, it is possible to trace people’s municipality of residence, determined at the end of each year. We treat the first municipality registered as the municipality of assignment, if the refugee is recorded in the registries the year in which asylum was granted, or on the subsequent year.18

The specific date in which the household head has been assigned to the municipality is key for our identification strategy considering the above-mentioned new findings. The date recorded in the residence permit information, provided by the DIS, is the date at which the individual is recorded in the municipal population register, which happens quite fast after the refugees receive their residence permit.19 With the exception of UN quota refugees (of which Denmark until 2016 invited 500 annually), applicants for

16 Leaving the panel is caused by out-migration, but only 1% of the gross sample out-migrate within the four-year observation period. 17 The household head is defined as the parent, or the father if both parents were granted asylum on the same date, at working age in the family, between 25 and 59 years old. The reason for choosing the male partner is that, as the data descriptive will show, in the context of non-western immigrants, men have a higher labour market attachment than women. 18 Refugees who are neither observed in the population registers in the year of asylum nor the subsequent year are excluded from our balanced sample of refugee household heads. 19 It takes on average 40 days from refugees are permitted residence until they are registered in the municipality population register in the period 2005-2010, for which the calculation has been made (Hvidtfeldt and Schultz-Nielsen 2017).

16 14

asylum apply after arrival to Denmark and live in a refugee reception centre until the decision on their application for asylum.

As shown in previous studies, the educational level of refugees before immigration can influence their integration into the labour market integration (e.g. Damm 2009). Information regarding educational attainment from abroad is generally obtained through surveys conducted by Statistics Denmark. In case of non-response Statistics Denmark imputes the value, but in order to avoid endogeneity issues, we have excluded this information and considered the educational level unknown in those cases. Information regarding subsequent education (obtained in Denmark) is also excluded as it may suffer from selection bias. However, using panel data estimation techniques allows us to control for time constant unobserved individual heterogeneity, e.g. innate abilities.

The administrative-territorial structure of Denmark has undergone a major structural reform from 2003 on that culminated in the local and regional government reform of 2007. Prior to the reform, the administrative division consisted of 14 counties and 271 municipalities. The reform abolished the counties and replaced them with five regions while reducing the number of municipalities to 98 (LGDK, 2009). However, not all the municipalities translated one-to-one to the new municipalities, disrupting the continuity of the dataset at the 2007 mark. This affects the municipality level variables: the municipality quotas, LLM and the municipality of placement of the refugees. Twelve municipalities are split into two and one municipality, Aalestrup, is split into three municipalities. We solve the data break by assigning the full population of these thirteen municipalities to the post-reform municipality to which the majority of the previous municipal population belonged post-municipal. Even though this does not reflect the reality perfectly, this higher order inaccuracy is expected to have a low impact on the later investigation as only 2% of the national population lives in these thirteen municipalities and only 3% of the individuals granted asylum are allocated in those municipalities.

Summing up, the sample selection criteria for the balanced panel is as follows. It is restricted to the subset of refugees in the gross sample who are aged 18-59, are observed in the administrative registers in four years after the year of asylum, and are household heads. These sample selection criteria for the balanced panel of household heads result in a dataset with observations for 8,479 household heads.20

The third dataset we use is a subsample of the balanced panel of household heads. As explained earlier we restrict this sample not only to individuals subject to the Danish Dispersal Policy, but also to those arriving after the first municipality quotas have been filled in order for refugees to be even less likely to influence municipality assignment.

20 For a detailed description of the sample reduction after each sample selection criteria see Table A1 in the appendix.

15

In principle, we could have restricted the sample to those refugees arriving in the very last month of the calendar year, but we would then have a very small sample and in years in which municipal quotas are never filled, their settlement may still be endogenous. Instead, we have investigated the random distribution of refugee household heads across municipalities by running balancing tests for subsamples of the balanced samples excluding household heads who receive asylum before the first municipality has filled its quota in that year, then excluding household heads who receive asylum before the first two, first three and so forth up to the first 10 municipalities have filled their quotas. Based on the results we have chosen to restrict our subsample of the balanced panel of household heads to household heads who get asylum after the first 10 municipalities within a given year have their quotas filled. This selection criterion reduces our subsample from the balanced panel to observations for 4,282 household heads.

Household heads are important as they represent the main subjects of the dispersal policy, whose settlement other family members’ settlement depends on. If household heads are distributed completely randomly across the country, we should expect there to be no correlation between personal characteristics and characteristics of the municipality of assignment. If on the other hand such correlations exist we need (as a minimum) to control for these possible correlated effects to identify the true relationship between local labour demand characteristics and individuals’ later employment.

For all samples, we merge the information on the first residence of all individuals with municipality time-series data constructed by the authors. Using primarily administrative registers from Statistics Denmark, we calculate the population share and non-Western immigrant share, co-national share and labour market characteristics for each municipality in the observation period.

In order to test if individuals in our sample were initially randomly distributed across municipalities with different labour demand characteristics in terms of educational attainment at asylum conditional on demographic characteristics known by DIS, we run balancing tests using the subsample of the balanced panel of household heads. The results are shown in Table 1.

[Table 1. Balancing tests. Include around here]

For each of the following characteristics of the municipality in the year of assignment: General unemployment rate, unemployment rate among non-Western immigrants, general employment rate, employment rate among non-Western immigrants, employment growth, population share, non-Western immigrants share, co-national share, three different measures of commuting distance to centre of local labour market, and the annual influx of assigned refugees per 1,000 inhabitants, we regress the municipality characteristic on the personal characteristics of the household head known by DIS at the time of the assignment that may therefore have influenced the municipal settlement decision and/or are important determinants of employment status 2-4 years after asylum:

17 14

asylum apply after arrival to Denmark and live in a refugee reception centre until the decision on their application for asylum.

As shown in previous studies, the educational level of refugees before immigration can influence their integration into the labour market integration (e.g. Damm 2009). Information regarding educational attainment from abroad is generally obtained through surveys conducted by Statistics Denmark. In case of non-response Statistics Denmark imputes the value, but in order to avoid endogeneity issues, we have excluded this information and considered the educational level unknown in those cases. Information regarding subsequent education (obtained in Denmark) is also excluded as it may suffer from selection bias. However, using panel data estimation techniques allows us to control for time constant unobserved individual heterogeneity, e.g. innate abilities.

The administrative-territorial structure of Denmark has undergone a major structural reform from 2003 on that culminated in the local and regional government reform of 2007. Prior to the reform, the administrative division consisted of 14 counties and 271 municipalities. The reform abolished the counties and replaced them with five regions while reducing the number of municipalities to 98 (LGDK, 2009). However, not all the municipalities translated one-to-one to the new municipalities, disrupting the continuity of the dataset at the 2007 mark. This affects the municipality level variables: the municipality quotas, LLM and the municipality of placement of the refugees. Twelve municipalities are split into two and one municipality, Aalestrup, is split into three municipalities. We solve the data break by assigning the full population of these thirteen municipalities to the post-reform municipality to which the majority of the previous municipal population belonged post-municipal. Even though this does not reflect the reality perfectly, this higher order inaccuracy is expected to have a low impact on the later investigation as only 2% of the national population lives in these thirteen municipalities and only 3% of the individuals granted asylum are allocated in those municipalities.

Summing up, the sample selection criteria for the balanced panel is as follows. It is restricted to the subset of refugees in the gross sample who are aged 18-59, are observed in the administrative registers in four years after the year of asylum, and are household heads. These sample selection criteria for the balanced panel of household heads result in a dataset with observations for 8,479 household heads.20

The third dataset we use is a subsample of the balanced panel of household heads. As explained earlier we restrict this sample not only to individuals subject to the Danish Dispersal Policy, but also to those arriving after the first municipality quotas have been filled in order for refugees to be even less likely to influence municipality assignment.

20 For a detailed description of the sample reduction after each sample selection criteria see Table A1 in the appendix.

15

In principle, we could have restricted the sample to those refugees arriving in the very last month of the calendar year, but we would then have a very small sample and in years in which municipal quotas are never filled, their settlement may still be endogenous. Instead, we have investigated the random distribution of refugee household heads across municipalities by running balancing tests for subsamples of the balanced samples excluding household heads who receive asylum before the first municipality has filled its quota in that year, then excluding household heads who receive asylum before the first two, first three and so forth up to the first 10 municipalities have filled their quotas. Based on the results we have chosen to restrict our subsample of the balanced panel of household heads to household heads who get asylum after the first 10 municipalities within a given year have their quotas filled. This selection criterion reduces our subsample from the balanced panel to observations for 4,282 household heads.

Household heads are important as they represent the main subjects of the dispersal policy, whose settlement other family members’ settlement depends on. If household heads are distributed completely randomly across the country, we should expect there to be no correlation between personal characteristics and characteristics of the municipality of assignment. If on the other hand such correlations exist we need (as a minimum) to control for these possible correlated effects to identify the true relationship between local labour demand characteristics and individuals’ later employment.

For all samples, we merge the information on the first residence of all individuals with municipality time-series data constructed by the authors. Using primarily administrative registers from Statistics Denmark, we calculate the population share and non-Western immigrant share, co-national share and labour market characteristics for each municipality in the observation period.

In order to test if individuals in our sample were initially randomly distributed across municipalities with different labour demand characteristics in terms of educational attainment at asylum conditional on demographic characteristics known by DIS, we run balancing tests using the subsample of the balanced panel of household heads. The results are shown in Table 1.

[Table 1. Balancing tests. Include around here]

For each of the following characteristics of the municipality in the year of assignment: General unemployment rate, unemployment rate among non-Western immigrants, general employment rate, employment rate among non-Western immigrants, employment growth, population share, non-Western immigrants share, co-national share, three different measures of commuting distance to centre of local labour market, and the annual influx of assigned refugees per 1,000 inhabitants, we regress the municipality characteristic on the personal characteristics of the household head known by DIS at the time of the assignment that may therefore have influenced the municipal settlement decision and/or are important determinants of employment status 2-4 years after asylum:

18 16

Years of education, which is our main variable of interest, but we also include age as well as indicators for being male, marital status, having children in different ages, country of origin, year and month of asylum.

The results reveal no signs on sorting in any of the municipality characteristics based on years of education. This is important because if refugees in different locations do not differ w.r.t. level of education at the time of asylum, they are also unlikely to differ in unobserved ways, e.g. w.r.t. to language proficiency. However, using a 5-percent significance level one or two demographic characteristics of the household head are correlated with a given municipality characteristics. Male household heads are less likely to be assigned to a municipality with a high unemployment rate and more likely to be assigned to a municipality with a high employment rate and longer distance to centre of local labour market, while there are no gender differences related to other municipality characteristics. Older household heads are more likely to be assigned to locations with a relatively high employment rate, longer distance to centre of local labour market and a higher annual influx of assigned refugees, while age of the household head is not correlated with other municipal characteristics. Married household heads are slightly less likely to be assigned to a location with a relatively high employment growth, while having children aged 3-17 is only slightly negatively correlated with the co-national share, and slightly positively correlated with the annual influx of assigned refugees. Given that we are testing 6 individual characteristics against 12 municipality characteristics it is not surprising that we find some correlations of which the gender-variation is the most common. We will continue to use all control variables in the analysis and will later conduct a robustness check where gender-specific estimates are presented.

According to Pei, Pischke and Schwandt (2017) a generally more powerful way of testing the relationship is to use the proxy for the candidate confounder (in our case educational level at the time of asylum) on the left-hand side of the regression instead of the right-hand side. Therefore, we have conducted this balancing test as well and shown it in Table A2. This test confirms that there is no correlation between individuals’ educational attainment (as measured by a dummy for having at least 10 years of education) and any of the 12 municipality characteristics.

Finally, we construct a dataset for analysis of the effects of local labour market conditions on the employment probability. The dataset consists of the subsample of the balanced panel augmented with observations for spouses who also get asylum on the same date as the household head (extracted from the gross sample of refugees); inclusion of such spouses augments the subsample of the balanced panel with observations for 814 individuals. These spouses were assigned to the same municipality as the household head at the same time. Inclusion of such spouses into our estimation sample increases efficiency of the estimations and increases the external validity of our results by inclusion

17

of more married female refugees into the sample. Henceforth, we refer to this sample as the subsample of household heads and jointly arrived couples; it has observations for 4,282 household heads and 814 spouses (arrived on the same date as the household head), summing to 5,096 individuals.

V.B. Summary statistics

Table 2 shows the summary statistics for our four samples: the gross sample of refugees, the balanced panel of household heads, the subsample of the balanced panel of household heads and the subsample of household heads and jointly arrived couples. The gross sample of refugees includes 8,400 men and 4,292 women summing up to 12,692 adult (18 years +) refugees that arrive for the first time to Denmark between 1999 and 2010. The balanced panel includes the 8,479 individuals from the gross sample that are household heads and observed in data during the first four years after their arrival. It is mainly the first criterion that reduces the sample. Recall that the subsample of the balanced panel of household heads includes those 4,282 individuals from the balanced panel of household heads that arrives after the first 10 municipality quotas are filled. Generally, refugees are often men travelling alone, while family reunified (arriving later) are more often women and children. Besides, in those cases where more family members arrive at the same date, we consider the man as household head and thereby only include him in the balanced sample and subsample. For these two reasons, men are overrepresented in both the balanced sample and the subsample of the balanced panel of household heads (82%) compared to the gross sample of refugees (66%). By contrast, the share of men in the subsample of household heads and jointly arrived couples (69%) is similar to the share of men in the gross sample of refugees.

[Table 2. Summary statistics of refugee sample. Include around here]

The employment rate of refugees increases by years since asylum for both genders, but it differs greatly between men and women. For men it increases from 32% in year 2 to 44% in year 4 since asylum. For women it increases from 10% in year 2 to 22% in year 4 since asylum.

A comparison of the individual characteristics in the ‘subsample of household heads and jointly arrived couples’ (our estimation sample) with the gross sample by gender shows that the exclusion of later arrived spouses as expected makes a larger difference for women than men.

The characteristics of the individual and the municipality of assignment summarized in Table 2 refer to the year of assignment. Generally, the individuals in the ‘subsample of household heads and jointly arrived spouses’ are a bit younger than in the gross sample (due to exclusion of refugees above age 60). The marriage rate is marginally higher in the subsample, while the share with children is also higher.

19 16

Years of education, which is our main variable of interest, but we also include age as well as indicators for being male, marital status, having children in different ages, country of origin, year and month of asylum.

The results reveal no signs on sorting in any of the municipality characteristics based on years of education. This is important because if refugees in different locations do not differ w.r.t. level of education at the time of asylum, they are also unlikely to differ in unobserved ways, e.g. w.r.t. to language proficiency. However, using a 5-percent significance level one or two demographic characteristics of the household head are correlated with a given municipality characteristics. Male household heads are less likely to be assigned to a municipality with a high unemployment rate and more likely to be assigned to a municipality with a high employment rate and longer distance to centre of local labour market, while there are no gender differences related to other municipality characteristics. Older household heads are more likely to be assigned to locations with a relatively high employment rate, longer distance to centre of local labour market and a higher annual influx of assigned refugees, while age of the household head is not correlated with other municipal characteristics. Married household heads are slightly less likely to be assigned to a location with a relatively high employment growth, while having children aged 3-17 is only slightly negatively correlated with the co-national share, and slightly positively correlated with the annual influx of assigned refugees. Given that we are testing 6 individual characteristics against 12 municipality characteristics it is not surprising that we find some correlations of which the gender-variation is the most common. We will continue to use all control variables in the analysis and will later conduct a robustness check where gender-specific estimates are presented.

According to Pei, Pischke and Schwandt (2017) a generally more powerful way of testing the relationship is to use the proxy for the candidate confounder (in our case educational level at the time of asylum) on the left-hand side of the regression instead of the right-hand side. Therefore, we have conducted this balancing test as well and shown it in Table A2. This test confirms that there is no correlation between individuals’ educational attainment (as measured by a dummy for having at least 10 years of education) and any of the 12 municipality characteristics.

Finally, we construct a dataset for analysis of the effects of local labour market conditions on the employment probability. The dataset consists of the subsample of the balanced panel augmented with observations for spouses who also get asylum on the same date as the household head (extracted from the gross sample of refugees); inclusion of such spouses augments the subsample of the balanced panel with observations for 814 individuals. These spouses were assigned to the same municipality as the household head at the same time. Inclusion of such spouses into our estimation sample increases efficiency of the estimations and increases the external validity of our results by inclusion

17

of more married female refugees into the sample. Henceforth, we refer to this sample as the subsample of household heads and jointly arrived couples; it has observations for 4,282 household heads and 814 spouses (arrived on the same date as the household head), summing to 5,096 individuals.

V.B. Summary statistics

Table 2 shows the summary statistics for our four samples: the gross sample of refugees, the balanced panel of household heads, the subsample of the balanced panel of household heads and the subsample of household heads and jointly arrived couples. The gross sample of refugees includes 8,400 men and 4,292 women summing up to 12,692 adult (18 years +) refugees that arrive for the first time to Denmark between 1999 and 2010. The balanced panel includes the 8,479 individuals from the gross sample that are household heads and observed in data during the first four years after their arrival. It is mainly the first criterion that reduces the sample. Recall that the subsample of the balanced panel of household heads includes those 4,282 individuals from the balanced panel of household heads that arrives after the first 10 municipality quotas are filled. Generally, refugees are often men travelling alone, while family reunified (arriving later) are more often women and children. Besides, in those cases where more family members arrive at the same date, we consider the man as household head and thereby only include him in the balanced sample and subsample. For these two reasons, men are overrepresented in both the balanced sample and the subsample of the balanced panel of household heads (82%) compared to the gross sample of refugees (66%). By contrast, the share of men in the subsample of household heads and jointly arrived couples (69%) is similar to the share of men in the gross sample of refugees.

[Table 2. Summary statistics of refugee sample. Include around here]

The employment rate of refugees increases by years since asylum for both genders, but it differs greatly between men and women. For men it increases from 32% in year 2 to 44% in year 4 since asylum. For women it increases from 10% in year 2 to 22% in year 4 since asylum.

A comparison of the individual characteristics in the ‘subsample of household heads and jointly arrived couples’ (our estimation sample) with the gross sample by gender shows that the exclusion of later arrived spouses as expected makes a larger difference for women than men.

The characteristics of the individual and the municipality of assignment summarized in Table 2 refer to the year of assignment. Generally, the individuals in the ‘subsample of household heads and jointly arrived spouses’ are a bit younger than in the gross sample (due to exclusion of refugees above age 60). The marriage rate is marginally higher in the subsample, while the share with children is also higher.

20 18

Information about immigrants’ education from abroad is less systematically gathered by Statistics Denmark between 2006 and 2016 which can explain why the share with unknown education varies by year of arrival. Apart from that the education distribution is quite stable across samples for men, while the share with ‘low education’ for women is smallest in the gross sample.

As the subsample consists of refugees arrived after the first municipality quotas have been filled and each year starts with new quotas in January, we find that none of the refugees in the subsample arrives in January/February, while this is equally common in the gross sample. As expected we also find a lower representation of individuals in the subsample arriving in years like 2002 and 2010, where the actual number of newly arrived refugees is lower than expected by the authorities the year before and equally higher representation of years like 1999 and 2001, where the opposite was the case. This can also explain why nationalities that arrive in the early years, like Afghans and Iraqis, have a bit higher representation in the subsample.

Regarding the municipality characteristics there is only little variation in the two samples and thereby not much sign that refugees should be able to self-select into more favourable municipalities. All in all, the main differences found between the gross sample and the subsample of household heads and jointly arrived spouses are closely related to the selection criterion for inclusion.

Summary statistics on local labour demand characteristics for the period 1999 to 2014 are reported in Table A3. The table includes 1568 observations corresponding to yearly information for each of the 98 municipalities over a period of 16 years. The local unemployment rate among all individuals aged 18 to 65 in Denmark is on average 4% in the municipalities, while the unemployment rate among non-Western immigrants is considerably higher, namely 12%. Unemployment rates are calculated as the share of unemployed in the work force. But not all individuals participate in the work force, some may be discouraged workers, who have given up searching for a job, some may not want a job. An alternative measure of local labour market conditions is therefore to use the employment rate calculated as the share of employed among all individuals in the work-age, which in this study is restricted to individuals aged 18-65. The average employment rate is 75% among all individuals, but considerably lower among non-Western immigrants of which only 50% are employed. The average local employment growth is 0. We also measure the share of the overall population in Denmark that is resident within each municipality. On average 1% of the total population is resident within a municipality. The share of non-Western immigrants is measured as their share of the total population within each municipality. On average 3% of the population in the municipalities are immigrants of non-Western origin.

The correlation between the different local labour demand characteristics is shown in Table A4. The result is that these labour demand measures are correlated, but also that

19

they are far from being equal to one, in which case there would be no reason to include more than one in the analysis. The measure of overall and non-Western local unemployment rate has a correlation of 0.59, meaning that municipalities with higher general unemployment rates also tend to have higher unemployment among non-Western immigrants. A reason for why the correlation is not even higher could be compositional differences among non-Western immigrants in different municipalities, e.g. country of origin or years since migration. The correlation between the general unemployment rate and employment rate is as expected close (-0.74), while the correlation between non-Western unemployment and employment rates within municipalities is not nearly as strong (-0.61), possibly due to municipality variation in labour force participation among non-Western immigrants. The correlation between the general unemployment rate and non-Western employment rate is as expected weaker (-0.44), but not as weak as the correlation between the general employment rate in municipality and the employment growth (0.31). The correlation between the general employment rate and the municipality characteristics ‘Population share’ (0.05) and ‘non-Western immigrant share’ (0.06) is even lower.

Table A5 provides detailed information about the definition of each variable and the data source used to construct each variable.

VI. RESULTS

VI.A. Baseline results

In this section we present our baseline results regarding the effects of local labour demand on immigrants’ employment status in Nov. of the year, 2-4 years since asylum. These estimations are based on the balanced sample of household heads and the subsample with jointly arrived couples subject to a spatial dispersal policy as described earlier. In accordance with Hoynes (2000) we present estimates for a wide range of labour market outcomes, but we extend the analysis since we not only report the OLS estimates but also include reduced form, 2SLS and first-stage estimates, following Damm (2014).

Results showing the coefficient estimate using five different local labour market measures are presented in Table 3. The five different measures are: unemployment rate (Panel A), unemployment rate among non-Western immigrants (Panel B), employment rate (Panel C), employment rate among non-Western immigrants (Panel D) and finally employment growth (Panel E). The dependent variable in all regressions is a dummy for being employed in November and the model is specified in accordance with the description in section IV and includes the following control variables: age, indicators for gender, marital status, children aged 0-2, children aged 3-17, educational attainment,

21 18

Information about immigrants’ education from abroad is less systematically gathered by Statistics Denmark between 2006 and 2016 which can explain why the share with unknown education varies by year of arrival. Apart from that the education distribution is quite stable across samples for men, while the share with ‘low education’ for women is smallest in the gross sample.

As the subsample consists of refugees arrived after the first municipality quotas have been filled and each year starts with new quotas in January, we find that none of the refugees in the subsample arrives in January/February, while this is equally common in the gross sample. As expected we also find a lower representation of individuals in the subsample arriving in years like 2002 and 2010, where the actual number of newly arrived refugees is lower than expected by the authorities the year before and equally higher representation of years like 1999 and 2001, where the opposite was the case. This can also explain why nationalities that arrive in the early years, like Afghans and Iraqis, have a bit higher representation in the subsample.

Regarding the municipality characteristics there is only little variation in the two samples and thereby not much sign that refugees should be able to self-select into more favourable municipalities. All in all, the main differences found between the gross sample and the subsample of household heads and jointly arrived spouses are closely related to the selection criterion for inclusion.

Summary statistics on local labour demand characteristics for the period 1999 to 2014 are reported in Table A3. The table includes 1568 observations corresponding to yearly information for each of the 98 municipalities over a period of 16 years. The local unemployment rate among all individuals aged 18 to 65 in Denmark is on average 4% in the municipalities, while the unemployment rate among non-Western immigrants is considerably higher, namely 12%. Unemployment rates are calculated as the share of unemployed in the work force. But not all individuals participate in the work force, some may be discouraged workers, who have given up searching for a job, some may not want a job. An alternative measure of local labour market conditions is therefore to use the employment rate calculated as the share of employed among all individuals in the work-age, which in this study is restricted to individuals aged 18-65. The average employment rate is 75% among all individuals, but considerably lower among non-Western immigrants of which only 50% are employed. The average local employment growth is 0. We also measure the share of the overall population in Denmark that is resident within each municipality. On average 1% of the total population is resident within a municipality. The share of non-Western immigrants is measured as their share of the total population within each municipality. On average 3% of the population in the municipalities are immigrants of non-Western origin.

The correlation between the different local labour demand characteristics is shown in Table A4. The result is that these labour demand measures are correlated, but also that

19

they are far from being equal to one, in which case there would be no reason to include more than one in the analysis. The measure of overall and non-Western local unemployment rate has a correlation of 0.59, meaning that municipalities with higher general unemployment rates also tend to have higher unemployment among non-Western immigrants. A reason for why the correlation is not even higher could be compositional differences among non-Western immigrants in different municipalities, e.g. country of origin or years since migration. The correlation between the general unemployment rate and employment rate is as expected close (-0.74), while the correlation between non-Western unemployment and employment rates within municipalities is not nearly as strong (-0.61), possibly due to municipality variation in labour force participation among non-Western immigrants. The correlation between the general unemployment rate and non-Western employment rate is as expected weaker (-0.44), but not as weak as the correlation between the general employment rate in municipality and the employment growth (0.31). The correlation between the general employment rate and the municipality characteristics ‘Population share’ (0.05) and ‘non-Western immigrant share’ (0.06) is even lower.

Table A5 provides detailed information about the definition of each variable and the data source used to construct each variable.

VI. RESULTS

VI.A. Baseline results

In this section we present our baseline results regarding the effects of local labour demand on immigrants’ employment status in Nov. of the year, 2-4 years since asylum. These estimations are based on the balanced sample of household heads and the subsample with jointly arrived couples subject to a spatial dispersal policy as described earlier. In accordance with Hoynes (2000) we present estimates for a wide range of labour market outcomes, but we extend the analysis since we not only report the OLS estimates but also include reduced form, 2SLS and first-stage estimates, following Damm (2014).

Results showing the coefficient estimate using five different local labour market measures are presented in Table 3. The five different measures are: unemployment rate (Panel A), unemployment rate among non-Western immigrants (Panel B), employment rate (Panel C), employment rate among non-Western immigrants (Panel D) and finally employment growth (Panel E). The dependent variable in all regressions is a dummy for being employed in November and the model is specified in accordance with the description in section IV and includes the following control variables: age, indicators for gender, marital status, children aged 0-2, children aged 3-17, educational attainment,

22 20

country of origin, year of immigration, month of immigration, years since migration as well as the municipality’s share of the total population in Denmark and the share of non-Western immigrants in municipality.

[Table 3. Baseline results. Include around here.]

The coefficient estimates shown in the first column result from pooled OLS estimation of Eq. (1) (see section IV) using the balanced panel of household heads. There are 30,573 observations in the balanced panel as each of the 8,479 household heads (mentioned in Table 2) and the 1,712 spouses are observed 3 times (in year 2, 3 and 4). Recall that this panel may include refugees who had their location wish fulfilled upon asylum. Hence, the second column presents results from pooled OLS of Eq. (1) using the subsample of household heads and jointly arrived couples who arrive after the first ten municipality quotas are filled. This sample has 15,288 observations corresponding to 5,096 individuals observed 3 times. The panel data structure of the subsample is explicitly used in the third column that presents the coefficient estimates from the reduced form model in Eq. (3), where local labour market outcomes are measured by the initial value in the municipality of assignment. Coefficient estimates shown in column four result from 2SLS estimation of the IV-model in Eq. (2) using the subsample of household heads and jointly arrived couples; recall that each contemporaneous local labour market characteristic is instrumented by the initial value of the local labour market characteristic in the municipality of assignment. Columns five and six report the corresponding coefficient of the instrument in the first stage and the t-test statistic of insignificance of the instrument. As we only have one instrument per endogenous variable, we cannot produce an over-identification test. In cases with just one excluded variable, we report the t-statistic for insignificance of the excluded variable in the first-stage regression; in these cases the F-statistic can be calculated as the squared value of the t-statistic. Using the usual “rule of thumb” that the set of instrumental variables is strong if the F-statistic for joint insignificance of exclusion restrictions is 10 or above, the t-statistics in Table 3 confirm that the instruments are strong for all five labour demand measures reported in Panels A-E.

Focusing on the estimated coefficient of unemployment rate in panel A, we find that the estimated effects are all highly significant and (as expected) negative. Our estimates suggest that an increase by 1 percentage point in the local unemployment rate reduces the individual employment probability by 1.4-1.8 percentage point depending on the sample and estimator. In Panel B the coefficient estimates of the local unemployment rate among non-Western immigrants are lower in magnitude than in Panel A, which is natural given the higher unemployment rate of non-Western immigrants than natives, but still highly significant.

Panel C reports the results for estimation of Eq. (1)-(3) using the local employment rate as the measure of local labour market demand. Across models, the estimate is highly

21

significant and (as expected) positive. Estimates based on reduced form and 2SLS estimates are rather close. According to the 2SLS estimates, an increase in the local employment rate by 1 percentage point increases individual employment probability by 0.6-0.7 percentage point (henceforth abbreviated pp.). This corresponds to an increase in the employment probability of 2.1% around the mean. Using the local employment rate among non-Western immigrants as our measure of local labour demand in Panel D results in somewhat lower estimates compared to Panel C, especially in the reduced form and 2SLS specifications. In return, the absolute values of the reduced form and 2SLS estimates of the local employment rate among non-Western immigrants are almost identical to the estimates of the unemployment rate of non-Western immigrants in Panel B, columns three and four.

The last local labour market condition for which we provide estimate effects is employment growth (Panel E). Unlike the other panels we only find significant estimates from pooled OLS estimation of Eq. (1), whereas the reduced form and IV-models yield insignificant effects, suggesting that local employment growth is not of significance for refugees’ employment.

Since the local employment growth has an insignificant effect on the individual employment status of refugees 2-4 years after asylum according to the RE and 2SLS estimates, we can rule out this measure as our preferred measure of local labour market conditions. Our preferred measure is instead the local employment rate, because it maximizes the explained variance in the reduced form regressions (see the overall R-squared values reported in Table 3), perhaps because it is more sensitive to the number of discouraged workers than the unemployment rate as argued in Section III.

VI.B. Robustness and heterogeneity analyses

In the following we conduct series of robustness checks and heterogeneity analyses. Our first robustness check investigates whether our estimated effects of local labour demand are robust to inclusion of other local characteristics, like job search networks and commuting costs. In other words, whether our baseline models adequately control for correlated effects. We include in turn the following additional controls: Commuting time (using public transportation and car), commuting distance, share of co-nationals and finally commuting area fixed effects. We use the definition of commuting areas by Statistics Denmark (2016). According to Statistics Denmark, there were 29 commuting areas in Denmark in 2014. Results of estimation of the IV-model in Eq. (2) using the subsample of household heads and jointly arrived couples are reported in Table 4, in columns 1-6 for the local unemployment rate and in columns 7-12 for the local employment rate. The estimates of both measures are very robust to the inclusion of share of co-nationals, commuting time and distance. Of the estimates of additional

23 20

country of origin, year of immigration, month of immigration, years since migration as well as the municipality’s share of the total population in Denmark and the share of non-Western immigrants in municipality.

[Table 3. Baseline results. Include around here.]

The coefficient estimates shown in the first column result from pooled OLS estimation of Eq. (1) (see section IV) using the balanced panel of household heads. There are 30,573 observations in the balanced panel as each of the 8,479 household heads (mentioned in Table 2) and the 1,712 spouses are observed 3 times (in year 2, 3 and 4). Recall that this panel may include refugees who had their location wish fulfilled upon asylum. Hence, the second column presents results from pooled OLS of Eq. (1) using the subsample of household heads and jointly arrived couples who arrive after the first ten municipality quotas are filled. This sample has 15,288 observations corresponding to 5,096 individuals observed 3 times. The panel data structure of the subsample is explicitly used in the third column that presents the coefficient estimates from the reduced form model in Eq. (3), where local labour market outcomes are measured by the initial value in the municipality of assignment. Coefficient estimates shown in column four result from 2SLS estimation of the IV-model in Eq. (2) using the subsample of household heads and jointly arrived couples; recall that each contemporaneous local labour market characteristic is instrumented by the initial value of the local labour market characteristic in the municipality of assignment. Columns five and six report the corresponding coefficient of the instrument in the first stage and the t-test statistic of insignificance of the instrument. As we only have one instrument per endogenous variable, we cannot produce an over-identification test. In cases with just one excluded variable, we report the t-statistic for insignificance of the excluded variable in the first-stage regression; in these cases the F-statistic can be calculated as the squared value of the t-statistic. Using the usual “rule of thumb” that the set of instrumental variables is strong if the F-statistic for joint insignificance of exclusion restrictions is 10 or above, the t-statistics in Table 3 confirm that the instruments are strong for all five labour demand measures reported in Panels A-E.

Focusing on the estimated coefficient of unemployment rate in panel A, we find that the estimated effects are all highly significant and (as expected) negative. Our estimates suggest that an increase by 1 percentage point in the local unemployment rate reduces the individual employment probability by 1.4-1.8 percentage point depending on the sample and estimator. In Panel B the coefficient estimates of the local unemployment rate among non-Western immigrants are lower in magnitude than in Panel A, which is natural given the higher unemployment rate of non-Western immigrants than natives, but still highly significant.

Panel C reports the results for estimation of Eq. (1)-(3) using the local employment rate as the measure of local labour market demand. Across models, the estimate is highly

21

significant and (as expected) positive. Estimates based on reduced form and 2SLS estimates are rather close. According to the 2SLS estimates, an increase in the local employment rate by 1 percentage point increases individual employment probability by 0.6-0.7 percentage point (henceforth abbreviated pp.). This corresponds to an increase in the employment probability of 2.1% around the mean. Using the local employment rate among non-Western immigrants as our measure of local labour demand in Panel D results in somewhat lower estimates compared to Panel C, especially in the reduced form and 2SLS specifications. In return, the absolute values of the reduced form and 2SLS estimates of the local employment rate among non-Western immigrants are almost identical to the estimates of the unemployment rate of non-Western immigrants in Panel B, columns three and four.

The last local labour market condition for which we provide estimate effects is employment growth (Panel E). Unlike the other panels we only find significant estimates from pooled OLS estimation of Eq. (1), whereas the reduced form and IV-models yield insignificant effects, suggesting that local employment growth is not of significance for refugees’ employment.

Since the local employment growth has an insignificant effect on the individual employment status of refugees 2-4 years after asylum according to the RE and 2SLS estimates, we can rule out this measure as our preferred measure of local labour market conditions. Our preferred measure is instead the local employment rate, because it maximizes the explained variance in the reduced form regressions (see the overall R-squared values reported in Table 3), perhaps because it is more sensitive to the number of discouraged workers than the unemployment rate as argued in Section III.

VI.B. Robustness and heterogeneity analyses

In the following we conduct series of robustness checks and heterogeneity analyses. Our first robustness check investigates whether our estimated effects of local labour demand are robust to inclusion of other local characteristics, like job search networks and commuting costs. In other words, whether our baseline models adequately control for correlated effects. We include in turn the following additional controls: Commuting time (using public transportation and car), commuting distance, share of co-nationals and finally commuting area fixed effects. We use the definition of commuting areas by Statistics Denmark (2016). According to Statistics Denmark, there were 29 commuting areas in Denmark in 2014. Results of estimation of the IV-model in Eq. (2) using the subsample of household heads and jointly arrived couples are reported in Table 4, in columns 1-6 for the local unemployment rate and in columns 7-12 for the local employment rate. The estimates of both measures are very robust to the inclusion of share of co-nationals, commuting time and distance. Of the estimates of additional

24 22

observed municipality characteristics, only the share of co-nationals has a significant effect at a conventional 5 percent significance level, however, the estimate is not robust.

Our results are slightly more mixed, when introducing commuting area fixed effects and thereby controlling for all variation between the 29 commuting areas in Denmark. The estimated effect of the local unemployment rate is halved in magnitude (-0.8 pp.) and insignificant, while the estimated effect of the local employment rate is less affected (0.5 pp.), but the estimate is now insignificant due to larger standard errors, which is not very surprising given the specification.21 We therefore continue to consider the local employment rate as the better measure of our local labour market conditions.22

[Include table 4 around here.]

One may argue that when moving from one municipality to another it affects all municipality characteristics and not just the local employment rate. To take account of this possible concern we introduce a 2SLS-specification, where the three characteristics of the area of assignment (employment rate, share of population, non-Western immigrant share) from our baseline specification are estimated using the same municipality characteristics at the time of assignment as three exclusion restrictions. The results of 2SLS estimation with three (instead of one) endogenous variables in Eq. (2) are reported in Table 5. The results show that the instruments are very strong and that the estimated effect of the local employment rate (0.7 pp.) is highly significant and close to the level reported in Table 3. Moreover, the estimated effect of the population share is positive and significant. Similarly, the non-Western immigrant share is negative, but insignificant. Both of these results contrast the results in Damm and Rosholm (2010) of a negative effect of the population size and immigrant share on the hazard rate into first job. Possible explanations include differences in model specifications (mixed proportional hazard model versus panel data model with random effects, different functional form of the explanatory variables) and sample period (1986-1998 versus 1999-2010). Our findings show that, ceteris paribus, employment of refugees who were spatially dispersed to locations over the 1999-2010-period was unaffected by whether they were assigned to a location with a high share of non-Western immigrants in the population or not, but positively affected by assignment to a larger municipality. A one percentage point

21 Åslund and Roth (2007) also mention that they have experimented with a specification including both regional and cohort/time fixed effects. They conclude that the differences in within-region variation in unemployment over time appear to be too small to identify the effects of initial conditions. 22 We have also investigated the idea of estimating the effects of the employment rate of co-nationals who are living in the municipality of assignment at the time at which individual i is assigned to the municipality using only within-municipality variation, that is, including municipality of assignment fixed effects into Eq. 2-3. Unfortunately, there is insufficient time-variation in the employment rate of co-nationals at the municipal level for that specification. As shown in Damm (2014) estimation of the effects of the local employment rate of co-nationals exploiting only within-municipality variation requires neighbourhood data that currently only exist for Denmark for the period 1985-2004 (see Damm and Schultz-Nielsen 2008). Moreover, such neighbourhood effects analysis is beyond the scope of our study.

23

increase in the municipality’s share of the population living in Denmark increases the employment probability of refugees 2-4 years after asylum by 1.7 percentage points (corresponding to 5.5%).

[Table 5. Include table around here.]

One could also argue that the local labour market conditions for low-skilled would be a relevant focus since many non-Western immigrants work in such positions. We therefore conduct a robustness check where the local employment rates and local unemployment rates are calculated solely based on low-skilled individuals in Denmark. The result of focusing on local labour market conditions for low-skilled are reported in Table 6.

[Table 6. Include table around here.]

The results show that the estimated employment effect of the local unemployment rate for low skilled is highly significant, but estimates (-0.8--1.0 pp.) are somewhat lower than in Table 3, reflecting the higher unemployment level among low-skilled than the general population. Similarly, when using employment rates for low-skilled as local labour market measure we find highly significant, but slightly lower (0.5-0.6 pp.) estimates. These results support our main findings and do not imply any change in model specifications.

Finally, we have conducted two heterogeneity analyses, the first is related to gender, the second to year of arrival.

[Table 7. Include table around here.]

Table 7 reports the results from an IV-model similar to Eq. 2, except that the model includes an interaction term between the measure of labour market condition and a dummy for female, hereby allowing the estimated effect of local labour market conditions to differ by gender; as measure of local labour conditions we use the local unemployment rate (Panel A) and the local employment rate (Panel B). Both IV-models are estimated using the subsample of household heads and jointly arrived couples including 3,519 men and 1,577 women. The main effect is significant in both models and the coefficient estimate is very similar to the baseline estimate in Table 3, whereas the estimated effect of the interaction term is insignificant in both models. This means that the estimated effect of the local unemployment rate (local employment rate) does not vary significantly by gender. Our results suggest that a one percentage point change in the local unemployment rate (local employment rate) is 1.9 percentage points (0.8 percentage points) on the employment probability of both male and female refugees 2-4 years after asylum. However, in percentages, a percentage point change in each variable is substantially larger for female than male refugees: 13% versus 5% for the local unemployment rate and 5.3% versus 2.1% for the local employment rate. The reason is that the employment rate of female refugees 2-4 years after asylum is only 15% compared to 38% for male refugees.

25 22

observed municipality characteristics, only the share of co-nationals has a significant effect at a conventional 5 percent significance level, however, the estimate is not robust.

Our results are slightly more mixed, when introducing commuting area fixed effects and thereby controlling for all variation between the 29 commuting areas in Denmark. The estimated effect of the local unemployment rate is halved in magnitude (-0.8 pp.) and insignificant, while the estimated effect of the local employment rate is less affected (0.5 pp.), but the estimate is now insignificant due to larger standard errors, which is not very surprising given the specification.21 We therefore continue to consider the local employment rate as the better measure of our local labour market conditions.22

[Include table 4 around here.]

One may argue that when moving from one municipality to another it affects all municipality characteristics and not just the local employment rate. To take account of this possible concern we introduce a 2SLS-specification, where the three characteristics of the area of assignment (employment rate, share of population, non-Western immigrant share) from our baseline specification are estimated using the same municipality characteristics at the time of assignment as three exclusion restrictions. The results of 2SLS estimation with three (instead of one) endogenous variables in Eq. (2) are reported in Table 5. The results show that the instruments are very strong and that the estimated effect of the local employment rate (0.7 pp.) is highly significant and close to the level reported in Table 3. Moreover, the estimated effect of the population share is positive and significant. Similarly, the non-Western immigrant share is negative, but insignificant. Both of these results contrast the results in Damm and Rosholm (2010) of a negative effect of the population size and immigrant share on the hazard rate into first job. Possible explanations include differences in model specifications (mixed proportional hazard model versus panel data model with random effects, different functional form of the explanatory variables) and sample period (1986-1998 versus 1999-2010). Our findings show that, ceteris paribus, employment of refugees who were spatially dispersed to locations over the 1999-2010-period was unaffected by whether they were assigned to a location with a high share of non-Western immigrants in the population or not, but positively affected by assignment to a larger municipality. A one percentage point

21 Åslund and Roth (2007) also mention that they have experimented with a specification including both regional and cohort/time fixed effects. They conclude that the differences in within-region variation in unemployment over time appear to be too small to identify the effects of initial conditions. 22 We have also investigated the idea of estimating the effects of the employment rate of co-nationals who are living in the municipality of assignment at the time at which individual i is assigned to the municipality using only within-municipality variation, that is, including municipality of assignment fixed effects into Eq. 2-3. Unfortunately, there is insufficient time-variation in the employment rate of co-nationals at the municipal level for that specification. As shown in Damm (2014) estimation of the effects of the local employment rate of co-nationals exploiting only within-municipality variation requires neighbourhood data that currently only exist for Denmark for the period 1985-2004 (see Damm and Schultz-Nielsen 2008). Moreover, such neighbourhood effects analysis is beyond the scope of our study.

23

increase in the municipality’s share of the population living in Denmark increases the employment probability of refugees 2-4 years after asylum by 1.7 percentage points (corresponding to 5.5%).

[Table 5. Include table around here.]

One could also argue that the local labour market conditions for low-skilled would be a relevant focus since many non-Western immigrants work in such positions. We therefore conduct a robustness check where the local employment rates and local unemployment rates are calculated solely based on low-skilled individuals in Denmark. The result of focusing on local labour market conditions for low-skilled are reported in Table 6.

[Table 6. Include table around here.]

The results show that the estimated employment effect of the local unemployment rate for low skilled is highly significant, but estimates (-0.8--1.0 pp.) are somewhat lower than in Table 3, reflecting the higher unemployment level among low-skilled than the general population. Similarly, when using employment rates for low-skilled as local labour market measure we find highly significant, but slightly lower (0.5-0.6 pp.) estimates. These results support our main findings and do not imply any change in model specifications.

Finally, we have conducted two heterogeneity analyses, the first is related to gender, the second to year of arrival.

[Table 7. Include table around here.]

Table 7 reports the results from an IV-model similar to Eq. 2, except that the model includes an interaction term between the measure of labour market condition and a dummy for female, hereby allowing the estimated effect of local labour market conditions to differ by gender; as measure of local labour conditions we use the local unemployment rate (Panel A) and the local employment rate (Panel B). Both IV-models are estimated using the subsample of household heads and jointly arrived couples including 3,519 men and 1,577 women. The main effect is significant in both models and the coefficient estimate is very similar to the baseline estimate in Table 3, whereas the estimated effect of the interaction term is insignificant in both models. This means that the estimated effect of the local unemployment rate (local employment rate) does not vary significantly by gender. Our results suggest that a one percentage point change in the local unemployment rate (local employment rate) is 1.9 percentage points (0.8 percentage points) on the employment probability of both male and female refugees 2-4 years after asylum. However, in percentages, a percentage point change in each variable is substantially larger for female than male refugees: 13% versus 5% for the local unemployment rate and 5.3% versus 2.1% for the local employment rate. The reason is that the employment rate of female refugees 2-4 years after asylum is only 15% compared to 38% for male refugees.

26 24

Instrument validity requires that the initial local labour market conditions only affect the current individual employment status through the current labour market conditions. Scarring effects would render our instrument invalid, but our reduced form estimates would still be valid. Therefore, it is reassuring to know that the reduced form and 2SLS estimates of the local labour market conditions as measured by local unemployment rate or the local employment rate are not significantly different (Table 3). The 2SLS estimate of the local employment rate is slightly larger in magnitude than the pooled OLS estimate for the subsample of household heads and jointly arrived couples (Table 3). In other words, the pooled OLS estimate of the local employment rate is downward biased, indicating negative self-selection of refugees into locations with high employment rates, but at a very modest level in view of the similarity between the pooled OLS and 2SLS estimates. In Table 8 we report the RE-estimates of the local unemployment and employment rate for individuals arriving before and after 2003. Our heterogeneity checks show that the similarity between the reduced form and 2SLS estimates (and pooled OLS estimates) is driven by the first four cohorts of refugees in our sample (1999-2002) that constitute 70% of our sample. We can rule out that it is due to cohort differences in the relocation rate from the municipality of assignment. Instead, we speculate that it is due to the low annual influx of new refugees arriving in later years of our observation period; the lower annual influx of assigned refugees may have made it substantially easier to integrate later cohorts into the local labour market than earlier cohorts, making employment of later cohorts insensitive to the local employment rate.23

[Table 8. Included around here.]

VII. DISCUSSION AND CONCLUSIONS

Our results show substantial gender differences in the speed of labour market integration of refugees who were subject to the Danish Spatial Dispersal Policy in place from 1999 until 2016. While the employment rate of male refugees increased during the first four years after asylum to around 44%, the employment rate of female refugees increased more slowly to reach only 22% four years after asylum.

Most importantly, our study provides quasi-experimental evidence that refugees who were subject to the Danish Spatial Dispersal Policy in place from 1999 until 2016, ceteris paribus, had a higher employment probability if they were assigned to a municipality with favourable local labour market conditions, as measured e.g. by the local

23 To test this hypothesis, ideally one could include the annual influx of assigned refugees and the interaction between the annual influx of assigned refugees and the employment rate in the municipality of assignment as additional explanatory variables of interest in Eq. 2 and 3. However, in our case there is insufficient municipal variation for that specification.

25

employment rate or the unemployment rate. The effect of the local labour demand is economically significant. Residence in a municipality with a one percentage point higher employment rate increases the employment rate of refugees by 0.6-0.7 percentage points (or 2.1%) within the first four years of their stay in Denmark. The estimated effect a one percentage point higher local employment rate increases the employment probability of male and female refugees by the same amount in percentage points, but since the employment rate among female refugees is substantially lower than that of male refugees two to four years after asylum, the estimated effect of a one percentage point increase in the local employment rate is larger for female than male refugees in percentage terms.

An auxiliary finding is that two-four years after asylum the employment probability of refugees, who were subject to the Danish Spatial Dispersal Policy in place from 1999 until 2016, ceteris paribus, was also positively affected by assignment to a larger municipality (as measured by the municipality’s share of the overall population in Denmark). In other words, ceteris paribus, refugees found jobs faster in larger municipalities, perhaps due to negative attitudes towards immigrants outside the larger cities as found by Dustmann, Vasiljeva and Damm (2018).

An additional auxiliary finding is, that two-four years after asylum the employment probability of refugees, who were subject to the Danish Spatial Dispersal Policy in place from 1999 until 2016, was unaffected by whether they were assigned to a municipality with a high or low concentration of non-Western immigrants.

Viewed together, our two auxiliary findings questions whether spatial dispersal of refugees in fact increases labour market integration of refugees in the short run. However, some cautionary notes are needed to show that our analyses are not sufficient to draw a firm conclusion on the economic effects of spatial dispersal of new refugees across locations. First, in the absence of a spatial dispersal policy on refugees, newly recognized refugees would most likely settle in the Copenhagen area or in one of the other large cities in the country.24 However, since Copenhagen Municipality had a municipal quota of refugees equal to zero throughout our observation period, our analyses do not shed light on whether refugees would have had better employment prospect if they had initially settled in Copenhagen Municipality instead of being assigned to a smaller municipality. Second, our main result and auxiliary findings are obtained in regressions without control for the annual local influx of refugees relative to the local population; one may speculate that a lower local influx of refugees relative to the local population (obtained through spatial dispersal of new arrivals of refugees) increases the employment probability of new refugees due to lower competition for the same local jobs; such a

24 Before the first spatial dispersal policy on refugees in Denmark was implemented, refugees tended to settle in the capital area or another large city in Denmark (see e,g. Damm, 2005; Damm and Dustmann, 2014).

27 24

Instrument validity requires that the initial local labour market conditions only affect the current individual employment status through the current labour market conditions. Scarring effects would render our instrument invalid, but our reduced form estimates would still be valid. Therefore, it is reassuring to know that the reduced form and 2SLS estimates of the local labour market conditions as measured by local unemployment rate or the local employment rate are not significantly different (Table 3). The 2SLS estimate of the local employment rate is slightly larger in magnitude than the pooled OLS estimate for the subsample of household heads and jointly arrived couples (Table 3). In other words, the pooled OLS estimate of the local employment rate is downward biased, indicating negative self-selection of refugees into locations with high employment rates, but at a very modest level in view of the similarity between the pooled OLS and 2SLS estimates. In Table 8 we report the RE-estimates of the local unemployment and employment rate for individuals arriving before and after 2003. Our heterogeneity checks show that the similarity between the reduced form and 2SLS estimates (and pooled OLS estimates) is driven by the first four cohorts of refugees in our sample (1999-2002) that constitute 70% of our sample. We can rule out that it is due to cohort differences in the relocation rate from the municipality of assignment. Instead, we speculate that it is due to the low annual influx of new refugees arriving in later years of our observation period; the lower annual influx of assigned refugees may have made it substantially easier to integrate later cohorts into the local labour market than earlier cohorts, making employment of later cohorts insensitive to the local employment rate.23

[Table 8. Included around here.]

VII. DISCUSSION AND CONCLUSIONS

Our results show substantial gender differences in the speed of labour market integration of refugees who were subject to the Danish Spatial Dispersal Policy in place from 1999 until 2016. While the employment rate of male refugees increased during the first four years after asylum to around 44%, the employment rate of female refugees increased more slowly to reach only 22% four years after asylum.

Most importantly, our study provides quasi-experimental evidence that refugees who were subject to the Danish Spatial Dispersal Policy in place from 1999 until 2016, ceteris paribus, had a higher employment probability if they were assigned to a municipality with favourable local labour market conditions, as measured e.g. by the local

23 To test this hypothesis, ideally one could include the annual influx of assigned refugees and the interaction between the annual influx of assigned refugees and the employment rate in the municipality of assignment as additional explanatory variables of interest in Eq. 2 and 3. However, in our case there is insufficient municipal variation for that specification.

25

employment rate or the unemployment rate. The effect of the local labour demand is economically significant. Residence in a municipality with a one percentage point higher employment rate increases the employment rate of refugees by 0.6-0.7 percentage points (or 2.1%) within the first four years of their stay in Denmark. The estimated effect a one percentage point higher local employment rate increases the employment probability of male and female refugees by the same amount in percentage points, but since the employment rate among female refugees is substantially lower than that of male refugees two to four years after asylum, the estimated effect of a one percentage point increase in the local employment rate is larger for female than male refugees in percentage terms.

An auxiliary finding is that two-four years after asylum the employment probability of refugees, who were subject to the Danish Spatial Dispersal Policy in place from 1999 until 2016, ceteris paribus, was also positively affected by assignment to a larger municipality (as measured by the municipality’s share of the overall population in Denmark). In other words, ceteris paribus, refugees found jobs faster in larger municipalities, perhaps due to negative attitudes towards immigrants outside the larger cities as found by Dustmann, Vasiljeva and Damm (2018).

An additional auxiliary finding is, that two-four years after asylum the employment probability of refugees, who were subject to the Danish Spatial Dispersal Policy in place from 1999 until 2016, was unaffected by whether they were assigned to a municipality with a high or low concentration of non-Western immigrants.

Viewed together, our two auxiliary findings questions whether spatial dispersal of refugees in fact increases labour market integration of refugees in the short run. However, some cautionary notes are needed to show that our analyses are not sufficient to draw a firm conclusion on the economic effects of spatial dispersal of new refugees across locations. First, in the absence of a spatial dispersal policy on refugees, newly recognized refugees would most likely settle in the Copenhagen area or in one of the other large cities in the country.24 However, since Copenhagen Municipality had a municipal quota of refugees equal to zero throughout our observation period, our analyses do not shed light on whether refugees would have had better employment prospect if they had initially settled in Copenhagen Municipality instead of being assigned to a smaller municipality. Second, our main result and auxiliary findings are obtained in regressions without control for the annual local influx of refugees relative to the local population; one may speculate that a lower local influx of refugees relative to the local population (obtained through spatial dispersal of new arrivals of refugees) increases the employment probability of new refugees due to lower competition for the same local jobs; such a

24 Before the first spatial dispersal policy on refugees in Denmark was implemented, refugees tended to settle in the capital area or another large city in Denmark (see e,g. Damm, 2005; Damm and Dustmann, 2014).

28 26

finding would support spatial dispersal policies of refugees. Third, even if spatial dispersal of refugees does not increase labour market integration of adult refugees in the short run, it may increase human capital acquisition and lower crime rates of children of refugees if refugees subjected to the spatial dispersal policy are assigned to locations with relatively low shares of youth criminals.25

Moreover, we conclude that the local employment rate is the better measure of local labour demand for immigrant workers for the following reasons. First, the effects of the local employment rate remain after inclusion of additional observed characteristics of the local labour market. Second, the coefficient estimate of the local employment rate only decreases slightly in magnitude after inclusion of commuting area fixed effects. A likely reason for the finding that the local employment rate is a better measure of the effects of local labour demand for immigrants than the unemployment rate is that a substantial share of immigrants are discouraged workers during recessions in which case the unemployment rate will underestimate the share of individuals without a job and would like to work. By contrast, the employment rate to some extent captures the number of discouraged workers.

Our findings of significant effects of the local labour market conditions support the reform of the Danish Spatial Dispersal Policy in July 2016. Since the reform, refugee settlement in a municipality with good employment prospects should be given important consideration.26

References

Algan, Y., C. Dustmann, A. Glitz and A. Manning (2010), “The Economic Situation of First and Second Generation Immigrants in France, Germany and the United Kingdom”, Economic Journal, 120(Feb.), F4-F30.

Angrist, J. (2001), “Estimation of Limited Dependent Variable Models with Dummy Endogenous Regressors”, Journal of Business and Economic Statistics, 19(1), 2-28.

Bauer, T.K., M. Lofstrom and K.F. Zimmermann (2000). Immigration Policy, Assimilation of Immigrants and Natives' Sentiments towards Immigrants: Evidence from 12 OECD-Countries. IZA Discussion Paper, Volume 187.

25 Using observations for children of refugees who were subject to the Danish Spatial Dispersal Policy in place from 1986-1998 and below the minimum age at criminal responsibility at arrival, Damm and Dustmann (2014) find that growing up in a neighbourhood with a relatively high youth crime conviction rate, ceteris paribus, increases males’ probability to be convicted of a youth crime. 26 Bekendtgørelse om boligplacering af flygtninge, BEK number 243, dated the 21st of March 2018, (or ”Boligbekendtgørelsen” for short), Section 12.

27

Berman, E., J. Bound and S. Machin (1998), “Implications of Skill-Biased Technological Change: International Evidence”, Quarterly Journal of Economics, 113(4), 1245-1279.

Beaman, L.A. (2012), “Social Networks and the Dynamics of Labour Market Outcomes: Evidence from Refugees Resettled in the U.S.”, Review of Economic Studies, 79, 128-161.

Bertrand, M. and S. Mullainathan (2004), “Are Emily and Greg More Employable than Lakisha and Jamal? A Field Experiment on Labour Market Discrimination”, American Economic Review, 94(4), 991-1013.

Borjas, G.J. (1985), “Assimilation, Changes in Cohort Quality and the earnings of immigrants”, Journal of Labor Economics 3(4), 463–489.

Borjas, G.J. (1995), “Assimilation, Changes in Cohort Quality revisited: what happened to earnings in the 1980s?” Journal of Labor Economics 13(2), 201–245.

Bratsberg, B., O. Raaum and K. Røed (2014), “Immigrants, Labour Market Performance and Social Insurance”, Economic Journal 124, 644–683.

Bratsberg, B., O. Raaum and K. Røed (2017), “Immigrant labor market integration across admission classes”, Nordic Economic Policy Review, 17-54.

Card, D. and J. E. DiNardo (2002), “Skill-Biased Technological Change and Rising Wage Inequality: Some Problems and Puzzles”, Journal of Labor Economics, 20(4), 733-783.

Carlsson, M. and D-O. Rooth, (2007), “Evidence of ethnic discrimination in the Swedish labor market using experimental data”, Labour Economics, 14(4), 716–729.

Chiswick, B.R. (1978), “The Effect of Americanization on the Earnings of Foreign-born Men. Journal of Political Economy, 86(5), 897-921.

Cortes, K.E. (2004), Are Refugees Different from Economic Immigrants? Some Empirical Evidence on the Heterogeneity of Immigrant Groups in the United States, Review of Economics and Statistics 86(2), 465–480.

Damm, A.P. (2005), “The Danish Dispersal Policy on Refugee Immigrants 1986-1998: A Natural Experiment?” Aarhus School of Business, Department of Economics WP 05-3.

Damm, Anna P. and Marie L. Schultz-Nielsen, (2008), “Danish Neighbourhoods: Construction and Relevance for Measurement of Residential Segregation”, Danish Journal of Economics (Nationaløkonomisk Tidsskrift), 146(3): 241-262

Damm, A.P. (2009), “Ethnic Enclaves and Immigrant Labor Market Outcomes: Quasi-Experimental Evidence”, Journal of Labor Economics, 27(2), 281-314.

29 26

finding would support spatial dispersal policies of refugees. Third, even if spatial dispersal of refugees does not increase labour market integration of adult refugees in the short run, it may increase human capital acquisition and lower crime rates of children of refugees if refugees subjected to the spatial dispersal policy are assigned to locations with relatively low shares of youth criminals.25

Moreover, we conclude that the local employment rate is the better measure of local labour demand for immigrant workers for the following reasons. First, the effects of the local employment rate remain after inclusion of additional observed characteristics of the local labour market. Second, the coefficient estimate of the local employment rate only decreases slightly in magnitude after inclusion of commuting area fixed effects. A likely reason for the finding that the local employment rate is a better measure of the effects of local labour demand for immigrants than the unemployment rate is that a substantial share of immigrants are discouraged workers during recessions in which case the unemployment rate will underestimate the share of individuals without a job and would like to work. By contrast, the employment rate to some extent captures the number of discouraged workers.

Our findings of significant effects of the local labour market conditions support the reform of the Danish Spatial Dispersal Policy in July 2016. Since the reform, refugee settlement in a municipality with good employment prospects should be given important consideration.26

References

Algan, Y., C. Dustmann, A. Glitz and A. Manning (2010), “The Economic Situation of First and Second Generation Immigrants in France, Germany and the United Kingdom”, Economic Journal, 120(Feb.), F4-F30.

Angrist, J. (2001), “Estimation of Limited Dependent Variable Models with Dummy Endogenous Regressors”, Journal of Business and Economic Statistics, 19(1), 2-28.

Bauer, T.K., M. Lofstrom and K.F. Zimmermann (2000). Immigration Policy, Assimilation of Immigrants and Natives' Sentiments towards Immigrants: Evidence from 12 OECD-Countries. IZA Discussion Paper, Volume 187.

25 Using observations for children of refugees who were subject to the Danish Spatial Dispersal Policy in place from 1986-1998 and below the minimum age at criminal responsibility at arrival, Damm and Dustmann (2014) find that growing up in a neighbourhood with a relatively high youth crime conviction rate, ceteris paribus, increases males’ probability to be convicted of a youth crime. 26 Bekendtgørelse om boligplacering af flygtninge, BEK number 243, dated the 21st of March 2018, (or ”Boligbekendtgørelsen” for short), Section 12.

27

Berman, E., J. Bound and S. Machin (1998), “Implications of Skill-Biased Technological Change: International Evidence”, Quarterly Journal of Economics, 113(4), 1245-1279.

Beaman, L.A. (2012), “Social Networks and the Dynamics of Labour Market Outcomes: Evidence from Refugees Resettled in the U.S.”, Review of Economic Studies, 79, 128-161.

Bertrand, M. and S. Mullainathan (2004), “Are Emily and Greg More Employable than Lakisha and Jamal? A Field Experiment on Labour Market Discrimination”, American Economic Review, 94(4), 991-1013.

Borjas, G.J. (1985), “Assimilation, Changes in Cohort Quality and the earnings of immigrants”, Journal of Labor Economics 3(4), 463–489.

Borjas, G.J. (1995), “Assimilation, Changes in Cohort Quality revisited: what happened to earnings in the 1980s?” Journal of Labor Economics 13(2), 201–245.

Bratsberg, B., O. Raaum and K. Røed (2014), “Immigrants, Labour Market Performance and Social Insurance”, Economic Journal 124, 644–683.

Bratsberg, B., O. Raaum and K. Røed (2017), “Immigrant labor market integration across admission classes”, Nordic Economic Policy Review, 17-54.

Card, D. and J. E. DiNardo (2002), “Skill-Biased Technological Change and Rising Wage Inequality: Some Problems and Puzzles”, Journal of Labor Economics, 20(4), 733-783.

Carlsson, M. and D-O. Rooth, (2007), “Evidence of ethnic discrimination in the Swedish labor market using experimental data”, Labour Economics, 14(4), 716–729.

Chiswick, B.R. (1978), “The Effect of Americanization on the Earnings of Foreign-born Men. Journal of Political Economy, 86(5), 897-921.

Cortes, K.E. (2004), Are Refugees Different from Economic Immigrants? Some Empirical Evidence on the Heterogeneity of Immigrant Groups in the United States, Review of Economics and Statistics 86(2), 465–480.

Damm, A.P. (2005), “The Danish Dispersal Policy on Refugee Immigrants 1986-1998: A Natural Experiment?” Aarhus School of Business, Department of Economics WP 05-3.

Damm, Anna P. and Marie L. Schultz-Nielsen, (2008), “Danish Neighbourhoods: Construction and Relevance for Measurement of Residential Segregation”, Danish Journal of Economics (Nationaløkonomisk Tidsskrift), 146(3): 241-262

Damm, A.P. (2009), “Ethnic Enclaves and Immigrant Labor Market Outcomes: Quasi-Experimental Evidence”, Journal of Labor Economics, 27(2), 281-314.

30 28

Damm, A.P. (2014), “Neighborhood Quality and Labor Market Outcomes: Evidence from Quasi-Random Neighborhood Assignment of Immigrants”, Journal of Urban Economics, 79, 139-166.

Damm, A.P. and C. Dustmann, (2014), “Does Growing Up in a High Crime Neighborhood affect Youth Criminal Behavior?” American Economic Review, 104(6): 1806-1832.

Damm, A.P. and M. Rosholm (2010), “Employment Effects of Spatial Dispersal of Refugees”, Review of Economics of the Household, 8(1): 105-146.

Dustmann, C. (1993), “Earnings Adjustment of Temporary Migrants.” Journal of Population Economics, 6, 153-168.

Dustmann, C., Glitz, A., and Vogel, T. (2010), Employment, Wages, and the Economic Cycle: Differences between Immigrants and Natives. European Economic Review, 54(1), 1–17.

Dustmann, C., and Görlach, J-P. (2015), Selective out-migration and the estimation of immigrants’ earnings profiles, in B.R. Chiswick and P.W. Miller (eds), Handbook of the Economics of International Migration, Volume 1A, 489–533, Elsevier.

Dustmann, C., K. Vasiljeva and A.P. Damm (2018), “Refugee Migration and Electoral Outcomes”, forthcoming in the Review of Economic Studies.

Edin, P-A, P. Fredriksson and O. Åslund (2003), “Ethnic Enclaves and the Economic Success of Immigrants - Evidence from a Natural Experiment”, Quarterly Journal of Economics, 118, 329-357.

Edin, P-A, P. Fredriksson and O. Åslund (2004), “Settlement policies and the economic success of immigrants”, Journal of Population Economics, 17(1), 133-155.

Esmail, A. and S. Everington (1997), “Asian Doctors Are Still Being Discriminated Against”, British Medical Journal, 314(May), 1619.

Goos, M., A. Manning and A. Salomons (2014), “Explaining Job Polarization: Routine-Biased Technological Change and Offshoring”, American Economic Review, 104(8), 2509-26.

Grossman, G.M. and E. Rossi-Hansberg (2008), “Trading Tasks: A Simple Theory of Offshoring”, American Economic Review, 98(5), 1978-1997.

Hoynes, H.W. (2000) “Local Labor Markets and Welfare Spells: Do Demand Conditions Matter? Review of Economics and Statistics, vol. 82(3), 351-368.

Husted, L., H.S. Nielsen, M. Rosholm and N. Smith (2001). Employment and Wage Assimilation of Male First-Generation Immigrants in Denmark. International Journal of Manpower, 22, 39–68.

29

Hvidtfeldt, C. and M.L. Schultz-Nielsen, (2017), “Flygtninge og asylansøgere i Danmark 1992-2016”, ROCKWOOL Foundation Research Unit, Working paper no. 50

Katz, L.F. and K.M. Murphy (1992), “Changes in Relative Wages, 1963-1987: Supply and Demand factors”, Quarterly Journal of Economics, 107(1), 35-78.

Kennan, J. (2013), “Open borders”, Review of Economic Dynamics, Vol. 16(2), L1–13.

Lalonde, R.J. and R.H. Topel, (1992) "The Assimilation of Immigrants in the U.S. Labor Market," In: Borjas, George J. and Richard B. Freeman (eds.) “Immigration and the Workforce: The Economic Consequences for the United States and Source Areas”. Chicago and London: University of Chicago Press, 67-92.

Lubotsky, D. (2007), “Chutes or Ladders? A Longitudinal Analysis of Immigrant Earnings”, Journal of Political Economy, 115

Moore, M.P. and P. Ranjan (2005), “Globalisation vs Skill-Biased Technological Change: Implications for Unemployment and Wage Inequality”, Economic Journal, 115(April), 391-422.

Munshi, K. (2003), “Networks in the Modern Economy: Mexican Migrants in the U.S. Labor Market”, Quarterly Journal of Economics, 118(2), 549-599.

Nielsen, C.P. and K.B. Jensen (2006), The Danish Integration Act’s Significance for the Settlement Patterns of Refugees, Copenhagen: Institute for Local Government Studies.

Pei, Z., J-S. Pischke, and H. Schwandt (2017), “Poorly Measured Confounders Are More Useful On the Left than on the Right”, NBER WP 23232.

Riach, P.A. and J. Rich (1991), “Testing for racial discrimination in the labour market”, Cambridge Journal of Economics, 15(3), 239-256.

Sarvimäki, M. (2011), Assimilation to a Welfare State: Labor Market Performance and Use of Social Benefits by Immigrants to Finland, Scandinavian Journal of Economics 113(3), 665–688.

Sarvimäkki, M. (2017), “Labour market integration of refugees in Finland”, Nordic Economic Policy Review, 91-114.

Schultz-Nielsen, M.L. (2016). Arbejdsmarkedstilknytningen for flygtninge og indvandrere. ROCKWOOL Foundation Research Unit. The University Press of Southern Denmark, Odense.

Schultz-Nielsen, M.L. (2017), “Labour market integration of refugees in Denmark”, Nordic Economic Policy Review, 55-90.

Smith, N., P.J. Pedersen, S. Pedersen, and M.L. Schultz-Nielsen, (2003). Fra mangel på arbejde til mangel på arbejdskraft. Copenhagen: Spektrum.

Statistics Denmark (2016). Færre og større pendlingsområder. DST Analyse 2016:23.

31 28

Damm, A.P. (2014), “Neighborhood Quality and Labor Market Outcomes: Evidence from Quasi-Random Neighborhood Assignment of Immigrants”, Journal of Urban Economics, 79, 139-166.

Damm, A.P. and C. Dustmann, (2014), “Does Growing Up in a High Crime Neighborhood affect Youth Criminal Behavior?” American Economic Review, 104(6): 1806-1832.

Damm, A.P. and M. Rosholm (2010), “Employment Effects of Spatial Dispersal of Refugees”, Review of Economics of the Household, 8(1): 105-146.

Dustmann, C. (1993), “Earnings Adjustment of Temporary Migrants.” Journal of Population Economics, 6, 153-168.

Dustmann, C., Glitz, A., and Vogel, T. (2010), Employment, Wages, and the Economic Cycle: Differences between Immigrants and Natives. European Economic Review, 54(1), 1–17.

Dustmann, C., and Görlach, J-P. (2015), Selective out-migration and the estimation of immigrants’ earnings profiles, in B.R. Chiswick and P.W. Miller (eds), Handbook of the Economics of International Migration, Volume 1A, 489–533, Elsevier.

Dustmann, C., K. Vasiljeva and A.P. Damm (2018), “Refugee Migration and Electoral Outcomes”, forthcoming in the Review of Economic Studies.

Edin, P-A, P. Fredriksson and O. Åslund (2003), “Ethnic Enclaves and the Economic Success of Immigrants - Evidence from a Natural Experiment”, Quarterly Journal of Economics, 118, 329-357.

Edin, P-A, P. Fredriksson and O. Åslund (2004), “Settlement policies and the economic success of immigrants”, Journal of Population Economics, 17(1), 133-155.

Esmail, A. and S. Everington (1997), “Asian Doctors Are Still Being Discriminated Against”, British Medical Journal, 314(May), 1619.

Goos, M., A. Manning and A. Salomons (2014), “Explaining Job Polarization: Routine-Biased Technological Change and Offshoring”, American Economic Review, 104(8), 2509-26.

Grossman, G.M. and E. Rossi-Hansberg (2008), “Trading Tasks: A Simple Theory of Offshoring”, American Economic Review, 98(5), 1978-1997.

Hoynes, H.W. (2000) “Local Labor Markets and Welfare Spells: Do Demand Conditions Matter? Review of Economics and Statistics, vol. 82(3), 351-368.

Husted, L., H.S. Nielsen, M. Rosholm and N. Smith (2001). Employment and Wage Assimilation of Male First-Generation Immigrants in Denmark. International Journal of Manpower, 22, 39–68.

29

Hvidtfeldt, C. and M.L. Schultz-Nielsen, (2017), “Flygtninge og asylansøgere i Danmark 1992-2016”, ROCKWOOL Foundation Research Unit, Working paper no. 50

Katz, L.F. and K.M. Murphy (1992), “Changes in Relative Wages, 1963-1987: Supply and Demand factors”, Quarterly Journal of Economics, 107(1), 35-78.

Kennan, J. (2013), “Open borders”, Review of Economic Dynamics, Vol. 16(2), L1–13.

Lalonde, R.J. and R.H. Topel, (1992) "The Assimilation of Immigrants in the U.S. Labor Market," In: Borjas, George J. and Richard B. Freeman (eds.) “Immigration and the Workforce: The Economic Consequences for the United States and Source Areas”. Chicago and London: University of Chicago Press, 67-92.

Lubotsky, D. (2007), “Chutes or Ladders? A Longitudinal Analysis of Immigrant Earnings”, Journal of Political Economy, 115

Moore, M.P. and P. Ranjan (2005), “Globalisation vs Skill-Biased Technological Change: Implications for Unemployment and Wage Inequality”, Economic Journal, 115(April), 391-422.

Munshi, K. (2003), “Networks in the Modern Economy: Mexican Migrants in the U.S. Labor Market”, Quarterly Journal of Economics, 118(2), 549-599.

Nielsen, C.P. and K.B. Jensen (2006), The Danish Integration Act’s Significance for the Settlement Patterns of Refugees, Copenhagen: Institute for Local Government Studies.

Pei, Z., J-S. Pischke, and H. Schwandt (2017), “Poorly Measured Confounders Are More Useful On the Left than on the Right”, NBER WP 23232.

Riach, P.A. and J. Rich (1991), “Testing for racial discrimination in the labour market”, Cambridge Journal of Economics, 15(3), 239-256.

Sarvimäki, M. (2011), Assimilation to a Welfare State: Labor Market Performance and Use of Social Benefits by Immigrants to Finland, Scandinavian Journal of Economics 113(3), 665–688.

Sarvimäkki, M. (2017), “Labour market integration of refugees in Finland”, Nordic Economic Policy Review, 91-114.

Schultz-Nielsen, M.L. (2016). Arbejdsmarkedstilknytningen for flygtninge og indvandrere. ROCKWOOL Foundation Research Unit. The University Press of Southern Denmark, Odense.

Schultz-Nielsen, M.L. (2017), “Labour market integration of refugees in Denmark”, Nordic Economic Policy Review, 55-90.

Smith, N., P.J. Pedersen, S. Pedersen, and M.L. Schultz-Nielsen, (2003). Fra mangel på arbejde til mangel på arbejdskraft. Copenhagen: Spektrum.

Statistics Denmark (2016). Færre og større pendlingsområder. DST Analyse 2016:23.

30

Storesletten, K. (2000), Sustaining fiscal policy through immigration, Journal of Political Economy 108, 300–23.

Åslund, O., A. Forslund, and L. Lijleberg (2017), “Labour market entry of non-labour migrants – Swedish evidence”, Nordic Economic Policy Review, 115-158.

Åslund, O., Hensvik, L. and Skans, O. N. (2014), Seeking similarity: How immigrants and natives manage in the labor market, Journal of Labor Economics, 32(3), 405-441.

Åslund, O., and D-O. Rooth (2007) “Do Where and When Matter? Initial Labour Market Conditions and Immigrant Earnings”, Economic Journal, 117(518), 422-448.

30

Storesletten, K. (2000), Sustaining fiscal policy through immigration, Journal of Political Economy 108, 300–23.

Åslund, O., A. Forslund, and L. Lijleberg (2017), “Labour market entry of non-labour migrants – Swedish evidence”, Nordic Economic Policy Review, 115-158.

Åslund, O., Hensvik, L. and Skans, O. N. (2014), Seeking similarity: How immigrants and natives manage in the labor market, Journal of Labor Economics, 32(3), 405-441.

Åslund, O., and D-O. Rooth (2007) “Do Where and When Matter? Initial Labour Market Conditions and Immigrant Earnings”, Economic Journal, 117(518), 422-448.

STATISTICS

34

12

34

56

7

1

0-12

yea

rs-0

.002

130.

130

0.13

00.

114

0.01

52-0

.014

90.

0400

(0.0

644)

(0.3

25)

(0.1

60)

(0.3

43)

(0.0

331)

(0.0

402)

(0.0

480)

M

ore

than

12

year

s-0

.042

9-0

.024

90.

231

0.12

9-0

.052

1-0

.034

20.

0770

(0.0

731)

(0.3

71)

(0.1

86)

(0.4

07)

(0.0

374)

(0.0

365)

(0.0

490)

Mal

e-0

.160

***

-0.5

64**

0.57

0***

0.68

4**

-0.0

240

-0.0

704*

0.00

778

(0.0

587)

(0.2

85)

(0.1

55)

(0.3

31)

(0.0

331)

(0.0

376)

(0.0

471)

Age

-0.0

0281

-0.0

0613

0.01

29*

0.02

350.

0015

0-0

.001

42-0

.003

44(0

.002

60)

(0.0

132)

(0.0

0698

)(0

.015

2)(0

.001

45)

(0.0

0190

)(0

.002

27)

Mar

ried

0.04

020.

0561

-0.1

79-0

.283

-0.0

679*

*-0

.027

9-0

.079

3*(0

.056

3)(0

.272

)(0

.148

)(0

.314

)(0

.031

4)(0

.038

9)(0

.046

4)Ch

ildre

n ag

ed 0

-2-0

.036

40.

273

0.15

60.

0615

-0.0

0470

0.01

60-0

.035

7(0

.069

3)(0

.323

)(0

.178

)(0

.360

)(0

.036

5)(0

.036

7)(0

.047

8)Ch

ildre

n ag

ed 3

-17

0.04

420.

0043

4-0

.057

80.

0915

-0.0

0144

-0.0

579*

-0.0

757*

(0.0

559)

(0.2

76)

(0.1

42)

(0.3

04)

(0.0

305)

(0.0

316)

(0.0

414)

Coun

try

of o

rigin

F.E

.Ye

sYe

sYe

sYe

sYe

sYe

sYe

sYe

ar o

f asy

lum

F.E

.Ye

sYe

sYe

sYe

sYe

sYe

sYe

sM

onth

of a

sylu

m F

.E.

Yes

Yes

Yes

Yes

Yes

Yes

Yes

Num

ber o

f ind

ivid

uals

4,28

24,

282

4,28

24,

282

4,28

24,

282

4,28

2R-

squa

red

0.34

30.

286

0.12

20.

351

0.71

20.

055

0.13

6

Popu

latio

n sh

are

Non-

Wes

tern

im

mig

rant

s sha

re

Tabl

e 1:

Ass

ignm

ent l

ocat

ion

attr

ibut

es a

nd in

divi

dual

char

acte

ristic

s of a

ssig

nees

(hou

seho

ld h

eads

).De

pend

ent v

aria

ble:

Cha

ract

erist

ics o

f mun

icipa

lity

of a

ssig

nmen

t in

year

of a

ssig

nmen

tUn

empl

oym

ent

rate

Unem

ploy

men

t ra

te o

f non

-W

este

rn

imm

igra

nts

Empl

oym

ent

rate

Empl

oym

ent

rate

of n

on-

Wes

tern

im

mig

rant

s

Empl

oym

ent

grow

th

Year

s of e

duca

tion

of h

ouse

hold

hea

d (re

f. ca

tego

ry 0

-9 y

ears

):

Note

: ***

: p<0

.01,

**:

p<0

.05,

*: p

<0.1

. Rob

ust s

tand

ard

erro

rs in

par

enth

eses

. Adm

inist

rativ

e re

gist

er in

form

atio

n fro

m S

tatis

tics D

enm

ark

from

199

7-20

15. T

he sa

mpl

e is

the

subs

ampl

e of

refu

gee

hous

ehol

d he

ads w

ho g

ot a

sylu

m d

urin

g 19

99-2

010,

who

wer

e ob

serv

ed in

the

first

four

yea

rs si

nce

asyl

um a

nd w

ho a

rriv

ed a

fter t

he fi

rst 1

0 m

unici

palit

ies h

ad b

een

fille

d in

the

year

of a

rriv

al. R

epor

ted

coef

ficie

nts a

re b

ased

on

linea

r reg

ress

ions

of m

unici

palit

y ch

arac

teris

tics i

n th

e ye

ar o

f ass

ignm

ent o

n in

divi

dual

ch

arac

teris

tics i

n th

e ye

ar o

f ass

ignm

ent.

Addi

tiona

l con

trol

: Dum

my

for m

issin

g in

form

atio

n on

edu

catio

nal a

ttai

nmen

t.

35

1011

12

1

0-12

yea

rs0.

594

-0.0

0550

-0.0

207

(0.6

69)

(0.0

0603

)(0

.017

7)

Mor

e th

an 1

2 ye

ars

-0.6

410.

0053

6-0

.033

3(0

.769

)(0

.006

03)

(0.0

211)

Mal

e1.

643*

*-0

.003

86-0

.004

87(0

.668

)(0

.006

76)

(0.0

171)

Age

0.04

730.

0001

280.

0019

6**

(0.0

310)

(0.0

0032

1)(0

.000

819)

Mar

ried

-0.8

45-0

.010

7*-0

.018

6(0

.630

)(0

.006

37)

(0.0

164)

Child

ren

aged

0-2

-0.9

680.

0102

0.01

43(0

.760

)(0

.006

55)

(0.0

189)

Child

ren

aged

3-1

70.

632

-0.0

157*

**0.

0364

**(0

.616

)(0

.005

71)

(0.0

166)

Coun

try

of o

rigin

F.E

.Ye

sYe

sYe

sYe

ar o

f asy

lum

F.E

.Ye

sYe

sYe

sM

onth

of a

sylu

m F

.E.

Yes

Yes

Yes

Num

ber o

f ind

ivid

uals

4,28

24,

282

4,28

2R-

squa

red

0.06

70.

430

0.37

4

0.09

36*

(0.0

530)

-1.5

09(1

.090

)

(1.1

25)

-0.9

67(1

.332

)1.

995*

(1.0

94)

Cha

ract

erist

ics o

f mun

icipa

lity

of a

ssig

nmen

t in

year

of a

ssig

nmen

t

Com

mut

ing

dist

ance

to ce

nter

of

loca

l lab

our m

arke

t by

car

9

0.75

0

Dist

ance

to

cent

er o

f loc

al

labo

r mar

ket

Co-n

atio

nal

shar

e

0.59

3

Note

: ***

: p<0

.01,

**:

p<0

.05,

*: p

<0.1

. Rob

ust s

tand

ard

erro

rs in

par

enth

eses

. Adm

inist

rativ

e re

gist

er in

form

atio

n fro

m S

tatis

tics D

enm

ark

from

199

7-20

15. T

he sa

mpl

e is

the

subs

ampl

e of

refu

gee

hous

ehol

d he

ads w

ho g

ot a

sylu

m d

urin

g 19

99-2

010,

who

wer

e ob

serv

ed in

the

first

four

yea

rs si

nce

asyl

um a

nd w

ho a

rriv

ed a

fter t

he fi

rst 1

0 m

unici

palit

ies h

ad b

een

fille

d in

the

year

of a

rriv

al. R

epor

ted

coef

ficie

nts a

re b

ased

on

linea

r reg

ress

ions

of m

unici

palit

y ch

arac

teris

tics i

n th

e ye

ar o

f ass

ignm

ent o

n in

divi

dual

ch

arac

teris

tics i

n th

e ye

ar o

f ass

ignm

ent.

Addi

tiona

l con

trol

: Dum

my

for m

issin

g in

form

atio

n on

edu

catio

nal a

ttai

nmen

t.

-0.6

57(1

.369

)1.

318

(1.0

66)

Yes

Yes

Yes

4,28

20.

075

Yes

4,28

20.

067

Yes

Yes

0.06

10*

(0.0

368)

-0.8

82(0

.747

)-1

.042

Year

s of e

duca

tion

of h

ouse

hold

hea

d (re

f. ca

tego

ry 0

-9 y

ears

):

(0.9

03)

0.62

9(0

.731

)

(0.7

93)

-0.9

07(0

.917

)1.

718*

*(0

.785

)

Annu

al in

flux o

f as

signe

d re

fuge

es

per 1

,000

in

habi

tant

s

Com

mut

ing

dist

ance

to ce

nter

of

loca

l lab

our m

arke

t by

publ

ic tr

ansp

orta

tion

8

Tabl

e 1

(con

tinue

d): A

ssig

nmen

t loc

atio

n at

trib

utes

and

indi

vidu

al ch

arac

teris

tics o

f ass

igne

es (h

ouse

hold

hea

ds).

Depe

nden

t var

iabl

e:

36

Men Women Men Women All Men Women AllOutcome variable:Employed 0.3880 0.1613 0.4017 0.1660 0.3341 0.3788 0.1484 0.3075

(0.4873) (0.3678) (0.4902) (0.3721) (0.4717) (0.4851) (0.3555) (0.4614)

By years since asylum:Employed in year 2 0.3196 0.101 0.3322 0.1044 0.2606 0.3018 0.0818 0.2337

(0.4663) (0.3014) (0.4710) (0.3059) (0.4302) (0.4591) (0.2741) (0.4232)

Employed in year 3 0.4021 0.1659 0.4149 0.1697 0.3524 0.3992 0.1535 0.3232

(0.4904) (0.3720) (0.4927) (0.3755) (0.4778) (0.4898) (0.3605) (0.4677)

Employed in year 4 0.4439 0.2186 0.4503 0.2239 0.3895 0.4354 0.2099 0.3656

(0.4969) (0.4134) (0.4982) (0.4170) (0.4877) (0.4959) (0.4074) (0.4816)

Background characteristics:Personal attributes:Male 1.00 0.00 1.00 0.00 0.82 1.00 0.00 0.69

(0.0000) (0.0000) (0.0000) (0.0000) (0.3827) (0.0000) (0.0000) (0.4623)Age at registration 33.49 34.55 32.59 33.18 32.66 32.58 32.47 32.54

(10.45) (12.33) (90.96) (10.80) (9.42) (9.05) (9.27) (9.12)Married 0.5796 0.6482 0.5967 0.3975 0,5631 0.6044 0.6639 0.6228

(0.4936) (0.4776) (0.4906) (0.4895) (0.4961) (0.4890) (0.4725) (0.4847)

0.1068 0.2395 0.1465 0.1031 0,1397 0.1469 0.2708 0.1852(0.3089) (0.4268) (0.3537) (0.3042) (0.3467) (0.3541) (0.4445) (0.3885)0.2294 0.4769 0.3980 0.4119 0,3921 0.3893 0.5910 0.4517

(0.4205) (0.4995) (0.4895) (0.4923) (0.4883) (0.4877) (0.4918) (0.4977)Educational attainment:Unknown education 0.3642 0.3204 0.3338 0.4067 0.298 0.2819 0.3519 0.3036

(0.4812) (0.4667) (0.4716) (0.4914) (0.4574) (0.4500) (0.4777) (0.4598)

Low education 0.2813 0.1545 0.2951 0.3460 0.3354 0.3242 0.3748 0.3399

(0.4497) (0.3614) (0.4561) (0.4758) (0.4722) (0.4682) (0.4842) (0.4737)

Medium education 0.2064 0.0785 0.2166 0.1540 0.2146 0.2299 0.1737 0.2125

(0.4048) (0.2690) (0.4120) (0.3611) (0.4106) (0.4208) (0.3790) (0.4091)

Higher education 0.1481 0.1040 0.1545 0.0933 0.152 0.1640 0.0996 0.1440

(0.3552) (0.3053) (0.3614) (0.2910) (0.3591) (0.3703) (0.2995) (0.3512)

Gross sample of refugees

Balanced panel of household heads

Subsample of balanced

panel of household

heads

Subsample of household heads and jointly arrived

couples

Table 2: Summary statistics. Mean (standard deviation).

Having a child aged 0-2

Having a child aged 3-17

37

Men Women Men Women All Men Women AllAsylum year:1999 0.1198 0.1139 0.1237 0.1312 0.2174 0.2137 0.1934 0.2074

(0.3247) (0.3178) (0.3292) (0.3377) (0.4125) (0.4100) (0.3951) (0.4055)

2000 0.1740 0.1533 0.1760 0.1580 0.1518 0.1577 0.1446 0.1536

(0.3792) (0.3603) (0.3809) (0.3648) (0.3589) (0.3645) (0.3518) (0.3606)

2001 0.2223 0.1885 0.2215 0.1991 0.2966 0.2987 0.2708 0.2900

(0.4158) (0.3911) (0.4153) (0.3994) (0.4568) (0.4577) (0.4445) (0.4538)

2002 0.1043 0.1260 0.1006 0.1436 0.0033 0.0026 0.0051 0.0033

(0.3056) (0.3319) (0.3008) (0.3508) (0.0571) (0.0505) (0.0711) (0.0577)

2003 0.0700 0.0801 0.0695 0.0698 0.068 0.0688 0.0881 0.0748

(0.2552) (0.2716) (0.2544) (0.2550) (0.2517) (0.2531) (0.2836) (0.2630)

2004 0.0317 0.0545 0.0304 0.0444 0.0241 0.0224 0.0342 0.0261

(0.1751) (0.2271) (0.1716) (0.2060) (0.1532) (0.1482) (0.1819) (0.1594)

2005 0.0308 0.0398 0.0315 0.0450 0.0381 0.0375 0.0399 0.0383

(0.1729) (0.1956) (0.1747) (0.2075) (0.1914) (0.1900) (0.1959) (0.1919)

2006 0.0356 0.0370 0.0373 0.0359 0.0383 0.0375 0.0412 0.0387

(0.1853) (0.1889) (0.1895) (0.1861) (0.1919) (0.1900) (0.1989) (0.1928)

2007 0.0363 0.0524 0.0337 0.0326 0.0329 0.0313 0.0520 0.0377

(0.1871) (0.2229) (0.1804) (0.1777) (0.1785) (0.1740) (0.2221) (0.1904)

2008 0.0431 0.0585 0.0430 0.0503 0.0556 0.0546 0.0736 0.0604

(0.2031) (0.2347) (0.2030) (0.2186) (0.2291) (0.2272) (0.2611) (0.2383)

2009 0.0464 0.0347 0.0465 0.0424 0.0654 0.0659 0.0488 0.0606

(0.2104) (0.1831) (0.2106) (0.2016) (0.2472) (0.2482) (0.2156) (0.2387)

2010 0.0857 0.0610 0.0862 0.0477 0.0086 0.0094 0.0082 0.0090

(0.2800) (0.2394) (0.2807) (0.2131) (0.0926) (0.0964) (0.0904) (0.0946)

Balanced panel of household heads

Subsample of balanced

panel of household

heads

Subsample of household heads and jointly arrived

couples

Table 2: Summary statistics. Mean (standard deviation). Continued.

Gross sample of refugees

38

Men Women Men Women All Men Women AllMonth of asylum:January 0.0679 0.0697 0.0662 0.0744 0.0000 0.0000 0.0000 0.0000

(0.2515) (0.2546) (0.2487) (0.2625) (0.0000) (0.0000) (0.0000) (0.0000)

February 0.0831 0.0881 0.0832 0.0829 0.0014 0.0011 0.0019 0.0014

(0.2760) (0.2834) (0.2762) (0.2758) (0.0374) (0.0337) (0.0436) (0.0370)

March 0.0771 0.0767 0.0759 0.0796 0.0255 0.0236 0.0266 0.0245

(0.2668) (0.2661) (0.2648) (0.2708) (0.1575) (0.1518) (0.1611) (0.1547)

April 0.0720 0.0750 0.0720 0.0731 0.0189 0.0190 0.0247 0.0208

(0.2585) (0.2635) (0.2585) (0.2604) (0.1362) (0.1367) (0.1554) (0.1427)

May 0.0855 0.0946 0.0838 0.0875 0.064 0.0637 0.0564 0.0614

(0.2796) (0.2927) (0.2771) (0.2826) (0.2448) (0.2442) (0.2308) (0.2401)

June 0.1337 0.1466 0.1382 0.1599 0.1448 0.1390 0.1598 0.1454

(0.3403) (0.3537) (0.3451) (0.3667) (0.3519) (0.3460) (0.3665) (0.3525)

July 0.1144 0.1100 0.1165 0.1038 0.146 0.1481 0.1566 0.1507

(0.3183) (0.3129) (0.3208) (0.3051) (0.3531) (0.3552) (0.3636) (0.3578)

August 0.0936 0.0913 0.0920 0.0992 0.1572 0.1552 0.1604 0.1568

(0.2912) (0.2881) (0.2890) (0.2991) (0.3640) (0.3621) (0.3671) (0.3636)

September 0.1185 0.1072 0.1193 0.1116 0.1836 0.1839 0.1737 0.1807

(0.3232) (0.3094) (0.3242) (0.3150) (0.3872) (0.3874) (0.3790) (0.3848)

October 0.1115 0.1014 0.1118 0.1012 0.1875 0.1898 0.1833 0.1878

(0.3148) (0.3018) (0.3152) (0.3017) (0.3904) (0.3922) (0.3870) (0.3906)

November 0.0375 0.0317 0.0361 0.0215 0.0612 0.0668 0.0457 0.0602

(0.1900) (0.1752) (0.1866) (0.1452) (0.2397) (0.2497) (0.2088) (0.2380)

December 0.0052 0.0079 0.0050 0.0052 0.01 0.0099 0.0108 0.0102

(0.0722) (0.0887) (0.0708) (0.0721) (0.0997) (0.0992) (0.1033) (0.1005)

Table 2: Summary statistics. Mean (standard deviation). Continued.

Gross sample of refugees

Balanced panel of household heads

Subsample of balanced

panel of household

heads

Subsample of household heads and jointly arrived

couples

39

Men Women Men Women All Men Women AllCountry of origin:Iraq 0.3113 0.1955 0.3065 0.2278 0.3286 0.3450 0.2055 0.3018

(0.4631) (0.3966) (0.4611) (0.4196) (0.4698) (0.4754) (0.4042) (0.4591)

Afghanistan 0.1901 0.1498 0.1994 0.1567 0.1969 0.2020 0.1617 0.1896

(0.3924) (0.3569) (0.3996) (0.3636) (0.3977) (0.4016) (0.3683) (0.3920)

Iran 0.0683 0.0524 0.0697 0.0542 0.0474 0.0492 0.0380 0.0457

(0.2523) (0.2229) (0.2546) (0.2264) (0.2125) (0.2162) (0.1914) (0.2089)

Somalia 0.0532 0.1139 0.0407 0.1717 0.0507 0.0296 0.0748 0.0436

(0.2245) (0.3178) (0.1977) (0.3772) (0.2194) (0.1694) (0.2632) (0.2041)

Syria 0.0355 0.0203 0.0373 0.0124 0.0215 0.0239 0.0146 0.0210

(0.1850) (0.1409) (0.1895) (0.1107) (0.1450) (0.1527) (0.1199) (0.1434)

Myanmar 0.0458 0.0431 0.0508 0.0209 0.0437 0.0492 0.0438 0.0475

(0.2091) (0.2031) (0.2196) (0.1431) (0.2044) (0.2162) (0.2046) (0.2127)

Yugoslavia 0.0406 0.0729 0.0417 0.0457 0.056 0.0546 0.1027 0.0695

(0.1974) (0.2600) (0.2000) (0.2089) (0.2300) (0.2272) (0.3037) (0.2543)

BosHz 0.0306 0.0615 0.0304 0.0379 0.0224 0.0227 0.0469 0.0302

(0.1722) (0.2403) (0.1716) (0.1909) (0.1481) (0.1491) (0.2115) (0.1712)

Serbia 0.0211 0.0438 0.0225 0.0248 0.0187 0.0190 0.0425 0.0263

(0.1436) (0.2047) (0.1482) (0.1556) (0.1354) (0.1367) (0.2018) (0.1600)

Russia 0.0207 0.0384 0.0199 0.0333 0.0189 0.0171 0.0374 0.0234

(0.1424) (0.1923) (0.1395) (0.1795) (0.1362) (0.1295) (0.1898) (0.1510)

0.1827 0.2083 0.1812 0.2148 0.1952 0.1878 0.2321 0.2015

(0.3865) (0.4061) (0.3852) (0.4108) (0.3964) (0.3906) (0.4223) (0.4012)

Subsample of balanced

panel of household

heads

Subsample of household heads and jointly arrived

couples

Table 2: Summary statistics. Mean (standard deviation). Continued.

Gross sample of refugees

Balanced panel of household heads

Country of origin <200 in sample

40

Men Women Men Women All Men Women AllMunicipality of assignment characteristics:Unemployment rate 4.1347 4.0513 4.1623 3.9613 4.1266 4.1696 4.0125 4.121

(1.5910) (1.5658) (1.6084) (1.4740) (1.6784) (1.7056) (1.6356) (1.6856)

13.9705 13.706 14.0881 13.4744 14.5838 14.7628 14.0451 14.5407

(7.4552) (7.0556) (7.5743) (6.9702) (7.8798) (8.0071) (7.1303) (7.7528)

Employment rate 76.5926 76.6762 76.5643 77.0716 76.8125 76.6978 76.9784 76.7846

(3.9109) (3.8295) (3.9057) (3.6579) (3.7759) (3.7972) (3.7628) (3.7885)

46.3182 46.461 46.2382 46.6662 46.0402 45.8603 46.667 46.1099

(8.7216) (8.7530) (8.7645) (8.3664) (9.1746) (9.2213) (9.1865) (9.2172)

Employment growth -0.1482 -0.0329 -0.1552 0.0730 0.3422 0.3254 0.3154 0.3223

(1.8772) (1.7481) (1.8797) (1.6223) (1.3785) (1.3893) (1.3696) (1.3831)

1.2052 1.2469 1.1699 1.2011 1.127 1.1288 1.0858 1.1155

(1.1334) (1.1589) (1.0359) (1.0748) (0.9206) (0.9021) (0.7992) (0.8717)

2.4453 2.541 2.3948 2.5234 2.2585 2.2366 2.2874 2.2524

(1.3132) (1.3023) (1.2295) (1.2307) (1.1523) (1.1415) (1.0602) (1.1171)

22.1335 21.7316 22.2939 23.6025 23.5481 23.2501 23.655 23.3754

(25.944) (25.200) (26.221) (24.667) (27.149) (27.640) (26.254) (27.217)

17.8545 17.299 17.8957 18.8973 18.7411 18.4587 18.8405 18.5768

(18.205) (17.991) (18.284) (17.988) (18.591) (18.711) (18.401) (18.615)

15.0672 14.6565 15.0893 16.0742 15.8048 15.5222 15.9779 15.6632

(15.377) (15.341) (15.419) (15.438) (15.665) (15.704) (15.668) (15.693)Co-national share 0.1723 0.1851 0.1639 0.1861 0.1440 0.1425 0.1401 0.1418

(0.2178) (0.2443) (0.2103) (0.2238) (0.1925) (0.1922) (0.1971) (0.1937)

0.6575 0.6157 0.6641 0.6236 0.7339 0.74 0.7088 0.7303

(0.4489) (0.4310) (0.4482) (0.4358) (0.4983) (0.4990) (0.4766) (0.4923)

8400 4292 6947 1532 4,282 3519 1577 5096

97 96 95 91 94 93 89 94

Table 2: Summary statistics. Mean (standard deviation). Continued.

Gross sample of refugees

Balanced panel of household heads

Subsample of balanced

panel of household

heads

Subsample of household heads and jointly arrived

couples

Note: Administrative register information from Statistics Denmark from 1997-2015. Except for the outcome variables, all characteristics refer to the year of assignment.

Unemployment rate of non-Western immigrants

Employment rate of non-Western immigrants

Population share

Non-Western immigrant share

Commuting distance to center of local labour market by public transportation

Commuting distance to center of local labour market by car

Distance to center of local labour market

Annual influx of assigned refugees per 1,000 inhabitants

Number of observationsNumber of municipalities

41

POLS Balanced Sample

POLS RE 2SLS2SLS First

Stage

t-test of insignificance

of the instrument

1 2 3 4 5 6Panel A:

-0.0172*** -0.0150*** -0.0135*** -0.0184*** 0.735*** 32.78(0.00324) (0.00362) (0.00425) (0.00588) (0.0224)

R2 0.166 0.180 0.179 0.180

Panel B:-0.00306*** -0.00298*** -0.00222*** -0.00465*** 0.478*** 12.36(0.000797) (0.000782) (0.000584) (0.00120) (0.0387)

R2 0.165 0.180 0.179 0.179Panel C:

0.00684*** 0.00556*** 0.00573*** 0.00647*** 0.886*** 80.43(0.00139) (0.00137) (0.00144) (0.00163) (0.0110)

R2 0.166 0.181 0.180 0.180Panel D:

0.00543*** 0.00558*** 0.00201*** 0.00362*** 0.559*** 18.72(0.000801) (0.000897) (0.000777) (0.00138) (0.0298)

R2 0.168 0.184 0.179 0.183Panel E:Employment growth 0.0151*** 0.0148*** -0.00432 -0.0207 0.209*** 6.13

(0.00235) (0.00318) (0.00843) (0.0400) (0.0341)

R2 0.166 0.180 0.178 0.170

Observations 30,573 15,288 15,288 15,288 15,288Number of individuals 5,096 5,096 5,096

Table 3: Effects of local labour demand. Dependent variable: Employed in Nov.

Unemployment Rate

Unemployment rate of non-Western immigrants

Employment rate

Employment rate of non-Western immigrants

Note: ***: p<0.01, **: p<0.05, *: p<0.1. Standard errors clustered by municipality of assignment in parentheses. Administrative register information from Statistics Denmark from 1997-2015. The sample is the subsample of refugee household heads who got asylum during 1999-2010, who were observed in the first four years since asylum and and who arrived after the first 10 municipalities had been filled in the year of arrival as well as jointly arrived spouses. Controls: Age, indicators for male, marital status, number of children aged 0-2 and 3-17, educational attainment, country of origin, year of asylum, month of asylum and years since asylum, as well as municipality of assignment characteristics (population share and non-Western immigrants share). Additional control in columns 3, 4 and 5: An individual random effect.

42

12

34

56

78

910

1112

Expl

anat

ory

varia

ble:

Cha

ract

erist

ic o

f mun

icip

ality

of a

ssig

nmen

t-0

.018

4***

-0.0

189*

**-0

.019

0***

-0.0

189*

**-0

.018

2***

-0.0

0832

(0.0

0468

)(0

.004

72)

(0.0

0477

)(0

.004

78)

(0.0

0468

)(0

.012

1)0.

0064

7***

0.00

650*

**0.

0066

2***

0.00

658*

**0.

0063

5***

0.00

469

(0.0

0139

)(0

.001

39)

(0.0

0141

)(0

.001

41)

(0.0

0139

)(0

.003

00)

Addi

tiona

l con

trol

s rel

ativ

e to

the

cont

rols

in T

able

3:

Com

mut

ing

time

usin

g pu

blic

tr

ansp

orta

tion

No

Yes

No

No

No

No

No

Yes

No

No

No

No

Com

mut

ing

time

by c

arN

oN

oYe

sN

oN

oN

oN

oN

oYe

sN

oN

oN

o

Com

mut

ing

dist

ance

No

No

No

Yes

No

No

No

No

No

Yes

No

No

Co-n

atio

nal

shar

eN

oN

oN

oN

oYe

sN

oN

oN

oN

oN

oYe

sN

o

Com

mut

ing

area

F.

E.N

oN

oN

oN

oN

oYe

sN

oN

oN

oN

oN

oYe

s

Num

ber o

f ob

serv

atio

nsN

umbe

r of

indi

vidu

als

Not

e: *

**: p

<0.0

1, *

*: p

<0.0

5, *

: p<0

.1. S

tand

ard

erro

rs c

lust

ered

by

mun

icip

ality

of a

ssig

nmen

t in

pare

nthe

ses.

Adm

inist

rativ

e re

gist

er in

form

atio

n fr

om S

tatis

tics D

enm

ark

from

199

7-20

15. T

he sa

mpl

e is

the

subs

ampl

e of

refu

gee

hous

ehol

d he

ads w

ho g

ot a

sylu

m d

urin

g 19

99-2

010,

who

wer

e ob

serv

ed in

the

first

four

yea

rs si

nce

asyl

um a

nd a

nd w

ho a

rriv

ed a

fter

the

first

10

mun

icip

aliti

es h

ad b

een

fille

d in

the

year

of a

rriv

al a

s wel

l as j

oint

ly a

rriv

ed sp

ouse

s. R

epor

ted

coef

ficie

nts a

re e

stim

ates

from

2SL

S es

timat

ion

with

an

indi

vidu

al ra

ndom

effe

ct.

Addi

tiona

l con

trol

s: A

ge, i

ndic

ator

s for

mal

e, m

arita

l sta

tus,

num

ber o

f chi

ldre

n ag

ed 0

-2 a

nd 3

-17,

edu

catio

nal a

ttai

nmen

t, co

untr

y of

orig

in, y

ear o

f asy

lum

, mon

th o

f asy

lum

and

yea

rs

since

asy

lum

, as w

ell a

s mun

icip

ality

of a

ssig

nmen

t cha

ract

erist

ics (

popu

latio

n sh

are

and

non-

Wes

tern

imm

igra

nts s

hare

).

Tabl

e 4:

Rob

ustn

ess c

heck

s. 2

SLS

estim

ates

.De

pend

ent v

aria

ble:

Em

ploy

ed in

Nov

.

15,2

88

5,09

6

Une

mpl

oym

ent

rate

Empl

oym

ent r

ate

43

2SLS First stage First stage First stage1 2 3 4

Panel A: Explanatory variables: Characteristics of the current municipality of residenceEmployment rate 0.00663***

(0.00147)Population share 0.0167**

(0.00807)-0.00260(0.00644)

Panel B: Instrumental variables: Characteristics of municipality of assignmentEmployment rate 0.886*** -0.00914** -0.00494

(0.00529) (0.00465) (0.00431)Population share -0.0619** 0.891*** -0.0288

(0.0280) (0.0246) (0.0228)0.111*** -0.0545*** 0.866***(0.0223) (0.0196) (0.0182)

Test of joint significance:

Number of observationsNumber of individuals

Table 5: Effects of characteristics of the current municipality of residence.

Employed in Nov. Employment rate in current municipality

of residence

Population share in current municipality

of residence

Non-Western immigrant share in

current municipality of residence

15,288

5,096

Note: ***: p<0.01, **: p<0.05, *: p<0.1. Standard errors clustered by municipality of assignment in parentheses. Administrative register information from Statistics Denmark from 1997-2015. The sample is the subsample of refugee household heads who got asylum during 1999-2010, who were observed in the first four years since asylum and and who arrived after the first 10 municipalities had been filled in the year of arrival as well as jointly arrived spouses. Reported coefficients in column 1 are estimates from 2SLS estimation with an individual random effect and reported coefficients in columns 2-4 are estimates from the first stage, also including an individual random effects. Additional controls: Age, indicators for male, marital status, number of children aged 0-2 and 3-17, educational attainment, country of origin, year of asylum, month of asylum and years since asylum, as well as municipality of assignment characteristics (population share and non-Western immigrants share).

F-test for joint insignificance of exclusion restrictions

Dependent variable:

Non-Western immigrants share

Non-Western immigrants share

F(3,4969) = 5324*** F(3,4969) = 470.9*** F(3,4969) = 812.5***

44

POLS Balanced Sample

POLS RE 2SLS2SLS First

Stage

t-test of insignificance of the instrument

1 2 3 4 5 6Panel A:

-0.0103*** -0.00921***-0.00772***-0.00976*** 0.791*** 24.38(0.00212) (0.00235) (0.00282) (0.00355) (0.0325)

Panel B:0.00610*** 0.00590*** 0.00534*** 0.00548*** 0.974*** 84.75

(0.00124) (0.00117) (0.00123) (0.00126) (0.0115)

Number of observations 30,573 15,288 15,288 15,288 15,288Number of individuals 5,096 5,096 5,096

Table 6: Robustness checks. Effects of local labour demand, low-skilled.

Unemployment rate of low skilled

Employment rate of low skilled

Note: ***: p<0.01, **: p<0.05, *: p<0.1. Standard errors clustered by municipality of assignment in parentheses. Administrative register information from Statistics Denmark from 1997-2015. The sample is the subsample of refugee household heads who got asylum during 1999-2010, who were observed in the first four years since asylum and and who arrived after the first 10 municipalities had been filled in the year of arrival as well as jointly arrived spouses. Reported coefficients are estimates from 2SLS estimation with an individual random effect. Controls: Age, indicators for male, marital status, number of children aged 0-2 and 3-17, educational attainment, country of origin, year of asylum, month of asylum and years since asylum, as well as municipality of assignment characteristics (population share and non-Western immigrants share). Low-skilled are defined as individuals having obtained primary and lower secondary education, but no further education.

Dependent variable: Employed in Nov.

45

(1) (2)

Panel A:Unemployment rate -0.0188**

(0.00879)Unemployment rate*dummy for female 0.00131

(0.0136)Panel B:Employment rate 0.00745***

(0.00227)Employment rate*dummy for female -0.00318

(0.00318)Number of observationsNumber of individuals

Table 7: Effects of local labour demand by gender. 2 SLS estimates.

Note: ***: p<0.01, **: p<0.05, *: p<0.1. Standard errors clustered by municipality of assignment in parentheses. Administrative register information from Statistics Denmark from 1997-2015. The sample is the subsample of refugee household heads who got asylum during 1999-2010, who were observed in the first four years since asylum and and who arrived after the first 10 municipalities had been filled in the year of arrival as well as jointly arrived spouses. Controls: Age, indicators for male, marital status, number of children aged 0-2 and 3-17, educational attainment, country of origin, year of asylum, month of asylum and years since asylum, as well as municipality of assignment characteristics (population share and non-Western immigrants share) and, finally, an individual random effect.

Dependent variable: Employed in Nov.

15,2885,096

46

1999-2002 2003-2010 1999-2002 2003-20101 2 3 4

Panel A: Characteristic of current municipality of residenceUnemployment rate -0.0251*** -21.17

(0.00526) (361.5)Employment rate 0.01000*** -0.00587

(0.00183) (0.00457)Panel B: Characteristic of assigned municipalityUnemployment rate -0.0193*** 0.0288***

(0.00503) (0.00791)Employment rate 0.00885*** -0.00385

(0.00171) (0.00331)Panel C: First-stage estimatesUnemployment rate 0.768*** -0.00130

(0.00860) (0.0213)Employment rate 0.884*** 0.656***

(0.00734) (0.0221)Controls:Age and dummies for male, marital status, number of children aged 0-2, number of children aged 3-17

Yes Yes Yes Yes

Educational attainment F.E. Yes Yes Yes YesCountry of origin F.E. Yes Yes Yes YesYear of asylum F.E. Yes Yes Yes YesMonth of asylum F.E. Yes Yes Yes YesAssigned municipality characteristics: Population share, non-Western immigrants share

Yes Yes Yes Yes

Number of observations 8,595 4,251 8,595 4,251Number of individuals 2,865 1,417 2,865 1,417

Immigration cohorts

Table 8: Effects of local labour demand by immigration cohort groups: 1999-2002 and 2003-2010. 2SLS estimates (Panel A) and reduced form estimates (Panel B). Dependent variable in

Panels A and B: Employed in Nov.

Note: ***: p<0.01, **: p<0.05, *: p<0.1. Standard errors clustered by municipality of assignment in parentheses. Administrative register information from Statistics Denmark from 1997-2015. The sample is the subsample of refugee household heads who got asylum during 1999-2010, who were observed in the first four years since asylum and and who arrived after the first 10 municipalities had been filled in the year of arrival as well as jointly arrived spouses. Additional control in Panels B and C: Individual random effect.

47

# Explanation Reduction Sample size

Panel 1 All the population in ophg1997-2016

Sample size after appending ophg files from 1997 to 20161,223,633

2 Drop all the observations without pnr (If pnr==.) 242,878 980,7553 Drop all the individuals imputed residence permit type in any given

yearIf an individual’s observation is imputed at any time

204,523 776,232

4 Keep only 1st residence permit of each individual 199,642 576,5905 Keep only the refugees 520,607 55,9836 Residence permit between 1999 and 2010 36,988 18,9957 Country of origin of the individual or spouse not Denmark 164 18,8318 The individual is found in the population register (BEF) at least

once between 1999 to 2016655 18,176

9Age at arrival is between 18 and 59, calculated as the year of receiving residence permit minus the date of birth recorded in BEF

5,764 12,412

10 First appearance in BEF is the same year or one after receiving residence permit , out of which:

197 12,215

Observed in year 0 63 Observed in year 1 12,152

11 Household head (below broken down) 3,192 9,023 Without partner (pnrp==.) 4,684 First refugee with a refugee partner arriving later (till<till_p) 2,512 Man with a refugee partner arriving together 1,827

12 Observed in RAS in years 2-4 544 8,479Sample for the Balancing test 8,479

Of which received residence permit after 10 municipalities' quotas were filled 4,282

Spouses uncontaminated subsampleRefugee spouses in working age getting asylum on the same date as the HHH that received residence permit once 10 municipalites were filled 821Of which observed in RAS in years 2-4 7 814

Total uncontaminated subsample 5,096

Table A1. Sample selection criteria.

48

1 2 3 4 5 6Explanatory variables: Characteristics of municipality of assignmentUnemployment rate -0.00151

(0.00475)0.000248(0.00101)

Employment rate 0.00211(0.00177)

0.000344(0.000882)

Employment growth -0.00308(0.00828)

Population share -0.00437(0.00652)

Controls:Man -0.0942*** -0.0938*** -0.0951*** -0.0942*** -0.0940*** -0.0943***

(0.0171) (0.0171) (0.0172) (0.0171) (0.0171) (0.0171)Age 0.00724*** 0.00724*** 0.00721*** 0.00723*** 0.00724*** 0.00723***

(0.000774) (0.000773) (0.000774) (0.000774) (0.000774) (0.000774)Married -0.0106 -0.0106 -0.0102 -0.0105 -0.0108 -0.0107

(0.0168) (0.0168) (0.0168) (0.0168) (0.0168) (0.0168)Children aged 0-2 0.0279 0.0279 0.0276 0.0279 0.0279 0.0280

(0.0193) (0.0193) (0.0193) (0.0193) (0.0193) (0.0193)Children aged 3-17 0.0147 0.0147 0.0148 0.0146 0.0147 0.0144

(0.0164) (0.0164) (0.0164) (0.0164) (0.0164) (0.0164)Country of origin F.E. Yes Yes Yes Yes Yes YesYear of immigration F.E. Yes Yes Yes Yes Yes YesMonth of immigration F.E. Yes Yes Yes Yes Yes YesNumber of observations 4,282 4,282 4,282 4,282 4,282 4,282R-squared 0.326 0.326 0.327 0.326 0.326 0.326Note: ***: p<0.01, **: p<0.05, *: p<0.1. Robust standard errors in parentheses. Administrative register information from Statistics Denmark from 1997-2015. The sample is the subsample of refugee households who got asylum during 1999-2010 and who were observed in the first four years since asylum. Reported coefficients are based on linear regressions of an indicator for having at least ten years of education at asylum on a characteristic of municipality of assignment and other individual characteristics in the year of assignment. Indicator for missing information on educational attainment included.

Table A2: Individual characteristics of assignees (household heads) and assignment location attributes.Dependent variable:

Household head has at least 10 years of education

Unemployment rate among Non-Western

Employment rate among Non-Western immigrants

49

7 8 9 10 11 12Explanatory variables: Characteristics of municipality of assignment

0.740(0.555)

2.00e-05(0.000233)

6.74e-06(0.000342)

8.01e-05(0.000405)

Co-national share -0.967(3.996)

Annual influx of assigned refugees -0.0257(0.0159)

Controls:Man -0.0940*** -0.0940*** -0.0940*** -0.0941*** -0.0940*** -0.0940***

(0.0171) (0.0171) (0.0171) (0.0171) (0.0171) (0.0171)Age 0.00726*** 0.00724*** 0.00724*** 0.00724*** 0.00724*** 0.00728***

(0.000774) (0.000774) (0.000774) (0.000774) (0.000774) (0.000773)Married -0.0100 -0.0106 -0.0106 -0.0106 -0.0107 -0.0111

(0.0168) (0.0168) (0.0168) (0.0168) (0.0168) (0.0168)Children aged 0-2 0.0282 0.0279 0.0279 0.0280 0.0280 0.0283

(0.0193) (0.0193) (0.0193) (0.0193) (0.0193) (0.0193)Children aged 3-17 0.0152 0.0146 0.0147 0.0146 0.0145 0.0156

(0.0164) (0.0164) (0.0164) (0.0164) (0.0164) (0.0164)Country of origin F.E. Yes Yes Yes Yes Yes YesYear of immigration F.E. Yes Yes Yes Yes Yes YesMonth of immigration F.E. Yes Yes Yes Yes Yes YesNumber of observations 4,282 4,282 4,282 4,282 4,282 4,282R-squared 0.327 0.326 0.326 0.326 0.326 0.327Note: ***: p<0.01, **: p<0.05, *: p<0.1. Robust standard errors in parentheses. Administrative register information from Statistics Denmark from 1997-2015. The sample is the subsample of refugee households who got asylum during 1999-2010 and who were observed in the first four years since asylum. Reported coefficients are based on linear regressions of an indicator for having at least ten years of education at asylum on a characteristic of municipality of assignment and other individual characteristics in the year of assignment. Indicator for missing information on educational attainment included.

Non-Western immigrant shareCommuting time to center of local labour market by Commuting time to center of local labour market by Distance to center of local labour market

Table A2 (continued): Individual characteristics of assignees (household heads) and assignment location attributes.

Dependent variable: Household head has at least 10 years of education

50

Mean (Std. Dev.)Unemployment rate 4.19

(1.69)Unemployment rate among Non-Western immigrants 12.13

(6.92)Employment rate 75.33

(4.31)49.7

(8.69)Employment growth -0.44

(1.88)Population (%) 1.02

(1.12)Immigrants (%) 5.63

(3.07)Non-Western immigrants (%) 3.45

(2.54)

Number of observations 1,568Note: Municipality characteristics are constructed using the administrative registers from Statistics Denmark for the period 1999-2014.

Employment rate among Non-Western immigrants

Table A3: Summary statistics of municipality characteristics over the 1999-2014-period.

51

Une

mpl

oym

ent r

ate

1.00

Une

mpl

oym

ent r

ate

of n

on-

Wes

tern

imm

igra

nts

0.59

1.00

Empl

oym

ent r

ate

-0.7

4-0

.28

1.00

Empl

oym

ent r

ate

of n

on-W

este

rn

imm

igra

nts

-0.4

4-0

.61

0.31

1.00

Empl

oym

ent g

row

th-0

.25

-0.0

60.

31-0

.02

1.00

Popu

latio

n sh

are

0.05

0.09

-0.1

6-0

.17

0.15

1.00

Imm

igra

nt sh

are

0.04

-0.0

8-0

.19

0.24

0.00

40.

331.

00N

on-W

este

rn im

mig

rant

shar

e0.

06-0

.01

-0.1

70.

220.

040.

270.

951.

00N

umbe

r of o

bser

vatio

ns1,

568

Empl

oym

ent

grow

th

Not

e: M

unic

ipal

ity c

hara

cter

istic

s are

con

stru

cted

usin

g th

e ad

min

istra

tive

regi

ster

s fro

m S

tatis

tics D

enm

ark

for t

he p

erio

d 19

99-2

014.

Tabl

e A4

: Cor

rela

tion

betw

een

mun

icip

ality

cha

ract

eris

tics i

n th

e 19

99-2

014-

perio

d.

Popu

latio

n sh

are

Imm

igra

nt

shar

e

Non

-Wes

tern

im

mig

rant

sh

are

Une

m-

ploy

men

t rat

e

Une

m-

ploy

men

t rat

e of

non

-W

este

rn

imm

igra

nts

Empl

oym

ent

rate

Empl

oym

ent

rate

of n

on-

Wes

tern

im

mig

rant

s

52

Variable Definition Primary data sourceRefugee Dummy for having the residence permit

type of refugee.Residence Permit Register (OPHG), Statistics Denmark (DST).

Date of residence permit Dates for residence permits imputed by the Immigration Service

Residence Permit Register (OPHG), Statistics Denmark (DST).

Household head Dummy for first-arrived adult in the household; if the spouses have arrived on the same date, the husband is defined as the household head.

Residence Permit Register (OPHG) and Population Register (BEF), Statistics Denmark (DST).

Municipality of assignment Municipality registered in the population registers in the year of receiving residence permit or the following year

Population register (BEF), DST.

Employed Dummy for being employed. Register-Based Labour Force Statistics (RAS), Statistics Denmark (DST).

Education level Education level before immigration, constructed based on an education code of the highest degree attained before immigration.

Survey-based register on immigrants' educational attainment before immigration, DST.

Country of origin Dummy for source country. Population register (BEF), DST.

Male Dummy for male. Population register (BEF), DST.

Age Age calculated as the observation year minus the year of birth observed in the population register

Population register (BEF), DST.

Married Dummy for married at arrival Population register (BEF), DST.

Child aged 0-2 Dummy for having a child aged 0-2 years. Population register (BEF), DST.

Child aged 3-17 Dummy for having a child aged 3-17 years.

Population register (BEF), DST.

Table A5.A: Variable definitions and primary data sources: Individual characteristics.

53

Variable Definition Primary data sourceMunicipality quota Annual maximum quota of refugees to

be allocated in the municipalityDanish Immigration Service (DIS)

Population share Number of inhabitants in the municipality divided by the total national population.

Population register, DST. Authors' calculations based on full population data.

Non-Western immigrants share Number of non-western immigrants living in the municipality divided by the number of inhabitants in the municipality.

Population register, DST. Authors' calculations based on full population data.

Co-national share Number of conationals living in the municipality divided by the number of inhabitants in the municipality.

Population register, DST. Authors' calculations based on full population data.

Unemployment rate Number of unemployed individuals in the municipality divided by the labour force of the municipality.

Population register and Register-Based Labour Force Statistics (RAS),DST. Authors' calculations based on full population data.

Unemployment rate of non-Western immigrants

Number of unemployed non-western immigrants in the municipality divided by the non-western immigrant labour force of the municipality.

Population register and Register-Based Labour Force Statistics (RAS),DST. Authors' calculations based on full population data.

Employment rate Number of employed individuals in the municipality divided by the number of individuals in the municipality in working age (18 to 65).

Population register and Register-Based Labour Force Statistics (RAS),DST. Authors' calculations based on full population data.

Employment rate of non-Western immigrants

Number of employed non-western immigrants in the municipality divided by the nubmer of non-western immigrants in working age.

Population register and Register-Based Labour Force Statistics (RAS),DST. Authors' calculations based on full population data.

Employment growth Percentage increase in the employed population compared to the previous year

Population register and Register-Based Labour Force Statistics (RAS),DST. Authors' calculations based on full population data.

Table A5.B Variable definitions and primary data sources: Area Characteristics.

Variable Definition Primary data source

Distance time by public transport

Time distance with public transportation from biggest station (either train or bus station) in the municipality to the central station of the communting area.

Calculated by using Google Maps. The time distance is calculated at Monday the 12th of March 2018 with arrival time 8 AM.

Distance time by car Time distance by car from the biggest station (either train or bus station) in the municipality to the central station of the communting area (the shortest distance in km).

Calculated by using Google Maps. The time distance is calculated at Monday the 12th of March 2018 with arrival time 8 AM.

Distance in kilometers Distance in kilometers Calculated by using Google Maps.

Communing area dummies Communtig areas are defined by Statistics Denmark. For a commuting (or Travel to Work) area it holds that i) the majority of the local employed population work in the area and that ii) the majority of the jobs in the area are occupied by people living in the area

Source: Statistics Denmark (2016), "Færre og større pendlingsområder". URL: https://www.dst.dk/da/Statistik/Analyser/visanalyse?cid=28054

Note: Administrative register information from Statistics Denmark for the years 1997-2015.

Table A5.C Variable definitions and primary data sources: Area Characteristics.